AINumbers Fintech Intelligence Suite

by ainumbers.co · remote · no sign-in to list tools · v0.4.26 · published by co.ainumbers

746 MCP tools across 1209 fintech tools: ChainGraph AP2 decisions, execution_hash. Zero PII.

Final · 12Changes · 390

Compare Registry entry

498 tools: 12 do something that can't be undone, 390 change things and 96 only read.

Scanned Oct 21, 2026

Deletes or cancels0Claude Code: deny
Can't be undone12Claude Code: ask: see these tools
Changes things390Claude Code: ask: see these tools
Only reads96Claude Code: allow: see these tools

93 checks matched something in the tool definitions. No finding is not a safety guarantee. How the checks work.

claude mcp add --transport http tools https://mcp.ainumbers.co/mcp

Findings

Matches of fixed checks, with the text they matched.

69 warn24 info

Warnbuild_workflow_links
The server marks "build_workflow_links" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "build_workflow_links" says "link": a word for a change.
What this check looks for
Warnrun_chain
The server marks "run_chain" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_chain" says "run": a word for a change.
What this check looks for
Warnrun_chain_batch
The server marks "run_chain_batch" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_chain_batch" says "run": a word for a change.
What this check looks for
Warncall_tool
The server marks "call_tool" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "call_tool" says "call": a word for a change.
What this check looks for
Warndecode_x402_payment
The server marks "decode_x402_payment" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "decode_x402_payment" says "payment": a word that usually means money moves, which cannot be taken back.
What this check looks for
Warnrun_kernel_vm
The server marks "run_kernel_vm" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_kernel_vm" says "run": a word for a change.
What this check looks for

Show 87 more findings

Warnchecklist_verify_run
The server marks "checklist_verify_run" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "checklist_verify_run" says "run": a word for a change.
What this check looks for
Warnintoto_record_chain_run
The server marks "intoto_record_chain_run" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":false,"openWorldHint":false}; The name "intoto_record_chain_run" says "run": a word for a change.
What this check looks for
Warnworkbook_evaluate
The server marks "workbook_evaluate" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "workbook_evaluate" says "evaluate": a word for a change.
What this check looks for
Warncheck_cash_leg_finality
The server marks "check_cash_leg_finality" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The input carries an amount (input.policy_parameters.amount) and a currency, so it moves money.
What this check looks for
Warnselect_agentic_checkout_protocol
The server marks "select_agentic_checkout_protocol" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "select_agentic_checkout_protocol" says "select": a word for a change.
What this check looks for
Warnrun_agentic_readiness_diagnostic
The server marks "run_agentic_readiness_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_agentic_readiness_diagnostic" says "run": a word for a change.
What this check looks for
Warnrun_mcp_deployability_diagnostic
The server marks "run_mcp_deployability_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_mcp_deployability_diagnostic" says "run": a word for a change.
What this check looks for
Warnrun_dora_readiness_diagnostic
The server marks "run_dora_readiness_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_dora_readiness_diagnostic" says "run": a word for a change.
What this check looks for
Warnrun_tempo_fit_diagnostic
The server marks "run_tempo_fit_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_tempo_fit_diagnostic" says "run": a word for a change.
What this check looks for
Warnmodel_tempo_payment_economics
The server marks "model_tempo_payment_economics" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "model_tempo_payment_economics" says "payment": a word that usually means money moves, which cannot be taken back.
What this check looks for
Warnscreen_tip20_transfer_batch
The server marks "screen_tip20_transfer_batch" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "screen_tip20_transfer_batch" says "transfer": a word for a change that another call does not undo.
What this check looks for
Warnrun_arc_fit_diagnostic
The server marks "run_arc_fit_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_arc_fit_diagnostic" says "run": a word for a change.
What this check looks for
Warnrun_treasury_clearing_fit
The server marks "run_treasury_clearing_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_treasury_clearing_fit" says "run": a word for a change.
What this check looks for
Warnrun_digital_trade_fit
The server marks "run_digital_trade_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_digital_trade_fit" says "run": a word for a change.
What this check looks for
Warnverify_trade_document_set
The server marks "verify_trade_document_set" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "verify_trade_document_set" says "set": a word for a change.
What this check looks for
Warnrun_tokenized_settlement_fit
The server marks "run_tokenized_settlement_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_tokenized_settlement_fit" says "run": a word for a change.
What this check looks for
Warnrun_agent_economy_fit
The server marks "run_agent_economy_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_agent_economy_fit" says "run": a word for a change.
What this check looks for
Warnverify_ap2_payment_receipt
The server marks "verify_ap2_payment_receipt" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "verify_ap2_payment_receipt" says "payment": a word that usually means money moves, which cannot be taken back.
What this check looks for
Warnrun_ai_act_highrisk_fit
The server marks "run_ai_act_highrisk_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_ai_act_highrisk_fit" says "run": a word for a change.
What this check looks for
Warnrun_carbon_compliance_fit
The server marks "run_carbon_compliance_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_carbon_compliance_fit" says "run": a word for a change.
What this check looks for
Warnapply_climate_scenario
The server marks "apply_climate_scenario" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "apply_climate_scenario" says "apply": a word for a change.
What this check looks for
Warnrun_t1_readiness_diagnostic
The server marks "run_t1_readiness_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_t1_readiness_diagnostic" says "run": a word for a change.
What this check looks for
Warnrun_pqc_timeline_fit
The server marks "run_pqc_timeline_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_pqc_timeline_fit" says "run": a word for a change.
What this check looks for
Warnrun_sanctions_screening_fit
The server marks "run_sanctions_screening_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_sanctions_screening_fit" says "run": a word for a change.
What this check looks for
Warnrun_mica_casp_fit
The server marks "run_mica_casp_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_mica_casp_fit" says "run": a word for a change.
What this check looks for
Warnlink_traceability_lot_code
The server marks "link_traceability_lot_code" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "link_traceability_lot_code" says "link": a word for a change.
What this check looks for
Warncrosswalk_agent_payment_rail_trust
The server marks "crosswalk_agent_payment_rail_trust" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "crosswalk_agent_payment_rail_trust" says "payment": a word that usually means money moves, which cannot be taken back.
What this check looks for
Warnassess_agent_directory_publish_readiness
The server marks "assess_agent_directory_publish_readiness" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "assess_agent_directory_publish_readiness" says "publish": a word for a change that another call does not undo.
What this check looks for
Warnvalidate_openvex_statement
The server marks "validate_openvex_statement" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "validate_openvex_statement" says "validate": a word for reading.
What this check looks for
Warnrun_emir_reporting_fit
The server marks "run_emir_reporting_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_emir_reporting_fit" says "run": a word for a change.
What this check looks for
Warnrun_vida_readiness_diagnostic
The server marks "run_vida_readiness_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_vida_readiness_diagnostic" says "run": a word for a change.
What this check looks for
Warnlink_eudr_supply_chain_traceability
The server marks "link_eudr_supply_chain_traceability" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "link_eudr_supply_chain_traceability" says "link": a word for a change.
What this check looks for
Warnrun_eudr_readiness_fit
The server marks "run_eudr_readiness_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_eudr_readiness_fit" says "run": a word for a change.
What this check looks for
Warnrun_ai_governance_fit
The server marks "run_ai_governance_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_ai_governance_fit" says "run": a word for a change.
What this check looks for
Warnrun_insurance_reporting_fit
The server marks "run_insurance_reporting_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_insurance_reporting_fit" says "run": a word for a change.
What this check looks for
Warnevaluate_irrbb_sot_eve
The server marks "evaluate_irrbb_sot_eve" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "evaluate_irrbb_sot_eve" says "evaluate": a word for a change.
What this check looks for
Warnevaluate_irrbb_sot_nii
The server marks "evaluate_irrbb_sot_nii" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "evaluate_irrbb_sot_nii" says "evaluate": a word for a change.
What this check looks for
Warnrun_irrbb_disclosure_fit
The server marks "run_irrbb_disclosure_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_irrbb_disclosure_fit" says "run": a word for a change.
What this check looks for
Warnconvert_markdown_document
The server marks "convert_markdown_document" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "convert_markdown_document" says "convert": a word for a change.
What this check looks for
Warnconvert_tabular_data
The server marks "convert_tabular_data" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "convert_tabular_data" says "convert": a word for a change.
What this check looks for
Warnselect_cbe_license
The server marks "select_cbe_license" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "select_cbe_license" says "select": a word for a change.
What this check looks for
Warngenerate_iscc_code
The server marks "generate_iscc_code" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "generate_iscc_code" says "generate": a word for a change.
What this check looks for
Warnselect_embedded_license
The server marks "select_embedded_license" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "select_embedded_license" says "select": a word for a change.
What this check looks for
Warngenerate_attribution_string
The server marks "generate_attribution_string" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "generate_attribution_string" says "generate": a word for a change.
What this check looks for
Warncheck_card_act_ability_to_pay
The server marks "check_card_act_ability_to_pay" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "check_card_act_ability_to_pay" says "pay": a word that usually means money moves, which cannot be taken back.
What this check looks for
Warnlint_lei_payment_binding
The server marks "lint_lei_payment_binding" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "lint_lei_payment_binding" says "payment": a word that usually means money moves, which cannot be taken back.
What this check looks for
Warncompute_parametric_trigger_payout
The server marks "compute_parametric_trigger_payout" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "compute_parametric_trigger_payout" says "trigger": a word for a change.
What this check looks for
Warnrun_illustration_selfsupport_test
The server marks "run_illustration_selfsupport_test" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_illustration_selfsupport_test" says "run": a word for a change.
What this check looks for
Warnvalidate_openids_homeowners_record
The server marks "validate_openids_homeowners_record" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "validate_openids_homeowners_record" says "validate": a word for reading.
What this check looks for
Warnscreen_onledger_transfer_batch
The server marks "screen_onledger_transfer_batch" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "screen_onledger_transfer_batch" says "transfer": a word for a change that another call does not undo.
What this check looks for
Warnrun_section125_ndt
The server marks "run_section125_ndt" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_section125_ndt" says "run": a word for a change.
What this check looks for
Warnrun_401k_adp_acp_test
The server marks "run_401k_adp_acp_test" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_401k_adp_acp_test" says "run": a word for a change.
What this check looks for
Warnrun_robinhood_chain_fit_diagnostic
The server marks "run_robinhood_chain_fit_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_robinhood_chain_fit_diagnostic" says "run": a word for a change.
What this check looks for
Warnbuild_amortization_schedule
The server marks "build_amortization_schedule" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "build_amortization_schedule" says "schedule": a word for a change.
What this check looks for
Warncompute_globe_topup_tax
The server marks "compute_globe_topup_tax" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "compute_globe_topup_tax" says "topup": a word that usually means money moves, which cannot be taken back.
What this check looks for
Warnrun_rate_shock_ladder
The server marks "run_rate_shock_ladder" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_rate_shock_ladder" says "run": a word for a change.
What this check looks for
Warnconvert_tempo_fee_amm
The server marks "convert_tempo_fee_amm" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "convert_tempo_fee_amm" says "convert": a word for a change.
What this check looks for
Warnrun_umr_aana_readiness
The server marks "run_umr_aana_readiness" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_umr_aana_readiness" says "run": a word for a change.
What this check looks for
Warnmap_call_report_schedule_rc
The server marks "map_call_report_schedule_rc" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "map_call_report_schedule_rc" says "call": a word for a change.
What this check looks for
Warnmap_call_report_schedule_rcr
The server marks "map_call_report_schedule_rcr" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "map_call_report_schedule_rcr" says "call": a word for a change.
What this check looks for
Warnrun_call_report_edit_checks
The server marks "run_call_report_edit_checks" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_call_report_edit_checks" says "run": a word for a change.
What this check looks for
Warnmap_bhc_schedule_hc
The server marks "map_bhc_schedule_hc" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "map_bhc_schedule_hc" says "schedule": a word for a change.
What this check looks for
Warnmap_bhc_schedule_hcr
The server marks "map_bhc_schedule_hcr" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "map_bhc_schedule_hcr" says "schedule": a word for a change.
What this check looks for
Infocall_tool
The description mentions "credential".
Matched text
description: "...only tools only: anything that issues a credential, stamps an anchor or reaches the networ..."
What this check looks for
Infocompare_agentic_payment_protocols
The description mentions "credential".
Matched text
description: "...Visa TAP, Mastercard Agent Pay) across credential, signing, scope, rail, identity, and au..."
What this check looks for
Infoexport_artifact
The description mentions "Credentials".
Matched text
description: "...BA taxonomy), and vc — a W3C Verifiable Credentials 2.0 rendering (OCG §13.11, application/..."
What this check looks for
Infoexport_artifact
The input schema at properties.format.description mentions "Credentials".
Matched text
inputSchema.properties.format.description: "...brl/vc implemented; vc = W3C Verifiable Credentials 2.0 (base profile, all nodes)."
What this check looks for
Infovc_issue
The title mentions "Credential".
Matched text
title: "Issue a signed W3C Verifiable Credential 2.0"
What this check looks for
Infovc_issue
The description mentions "Credential".
Matched text
description: "Composes and signs a W3C Verifiable Credential 2.0 with an eddsa-jcs-2022 Data Integri..."
What this check looks for
Infovc_issue
The input schema at properties.subject_id.description mentions "Credential".
Matched text
inputSchema.properties.subject_id.description: "Credential subject id (DID or any identifier strin..."
What this check looks for
Infoacdc_said_check
The title mentions "credential".
Matched text
title: "Recompute ACDC / vLEI credential SAIDs (structural integrity)"
What this check looks for
Infoacdc_said_check
The description mentions "credential".
Matched text
description: "...tifiers (SAIDs) of a pasted ACDC / vLEI credential JSON (top-level "d" plus any nested a/e..."
What this check looks for
Infoacdc_said_check
The input schema at properties.credential.description mentions "credential".
Matched text
inputSchema.properties.credential.description: "ACDC / vLEI credential JSON object (v, d, i, ri, s, a, e, r fi..."
What this check looks for
Infoacdc_said_check
The input schema at required[0] mentions "credential".
Matched text
inputSchema.required[0]: "credential"
What this check looks for
Infocheck_agent_attestation
The description mentions "credential".
Matched text
description: "...nd-policy-simulator, art-13-eudi-wallet-credential-readiness-checker. Output feeds: art-02..."
What this check looks for
Infocheck_agent_attestation
The input schema at properties.policy_parameters.description mentions "credential".
Matched text
inputSchema.properties.policy_parameters.description: "...ompute is "auto" or "server". Required: credential, validate_at_unix. Unknown keys are ref..."
What this check looks for
Infocheck_eudi_readiness
The title mentions "Credential".
Matched text
title: "EUDI Wallet Credential-Acceptance Readiness Checker"
What this check looks for
Infocheck_eudi_readiness
The description mentions "Credential".
Matched text
description: "EUDI Wallet Credential-Acceptance Readiness Checker: OpenChain..."
What this check looks for
Infovalidate_c2pa_manifest
The title mentions "Credential".
Matched text
title: "C2PA Content Credential Manifest Validator"
What this check looks for
Infovalidate_c2pa_manifest
The description mentions "Credential".
Matched text
description: "C2PA Content Credential Manifest Validator: OpenChainGraph comp..."
What this check looks for
Infoverify_content_credential_signature
The title mentions "Credential".
Matched text
title: "Content Credential Signature Verifier"
What this check looks for
Infoverify_content_credential_signature
The description mentions "Credential".
Matched text
description: "Content Credential Signature Verifier: OpenChainGraph comp..."
What this check looks for
Inforesolve_provenance_ingredient_tree
The description mentions "credential".
Matched text
description: "...pstream artifacts from: art-124-content-credential-signature-verifier. Open at: https://ai..."
What this check looks for
Infocheck_genius_reserve_disclosure
The description is 2,091 characters long, over the 2,000 this check looks for.
Matched text
description ends: "...serves_usd. Output schema: call describe_tool("check_genius_reserve_disclosure")."
What this check looks for
Infoverify_acdc_delegation_chain
The description mentions "credential".
Matched text
description: "...tic Chained Data Containers (ACDC): per-credential SAID self-addressing integrity, issuer-..."
What this check looks for
Infoverify_acdc_delegation_chain
The input schema at properties.policy_parameters.description mentions "credentials".
Matched text
inputSchema.properties.policy_parameters.description: "...ompute is "auto" or "server". Required: credentials, max_chain_depth. Unknown keys are refu..."
What this check looks for
Infoverify_revocation_status
The description mentions "credential".
Matched text
description: "...rom a supplied, zero-egress status list credential. Outputs: note, revoked_for_purpose, st..."
What this check looks for

Tools

Open a tool for its reasons and full description. Descriptions are the publisher's own words, shown as they are.

498 tools

Can't be undone 12Sends, pays, publishes or merges: once done, it cannot be taken back.Claude Code: ask

assess_agent_directory_publish_readinessAgent Directory Publish Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "assess_agent_directory_publish_readiness" says "publish": a word for a change that another call does not undo.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "assess_agent_directory_publish_readiness" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "assess_agent_directory_publish_readiness" says "publish": a word for a change that another call does not undo.
  What this check looks for
Description, as the server wrote it
Agent Directory Publish Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-133-agent-payment-rail-trust-crosswalk. Open at: https://ainumbers.co/chaingraph/art-134-agent-directory-publish-readiness.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_card_act_ability_to_payCheck CARD Act Ability to Pay: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "check_card_act_ability_to_pay" says "pay": a word that usually means money moves, which cannot be taken back.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "check_card_act_ability_to_pay" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "check_card_act_ability_to_pay" says "pay": a word that usually means money moves, which cannot be taken back.
  What this check looks for
Description, as the server wrote it
Check CARD Act Ability to Pay: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-228-build-adverse-action-notice. Open at: https://ainumbers.co/chaingraph/art-233-check-card-act-ability-to-pay.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Evaluates a credit card application against CARD Act §1026.51 ability-to-pay requirements. Outputs: ability_to_pay_result, annual_income, annual_minimum_payments_est, dti_ratio, dti_threshold, has_cosigner, income_and_assets, method_a_sufficient, method_b_sufficient, method_used, monthly_debt_obligations, monthly_housing_payment, monthly_income, monthly_minimum_payment_est, penalty_fee_safe_harbor, pii_note, regulatory_basis, requested_credit_limit, requires_cosigner, table_source, table_version, total_assets, total_monthly_obligations, under_21_restriction. Output schema: call describe_tool("check_card_act_ability_to_pay").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_cash_leg_finalityOn-Chain Cash-Leg Finality Checker: OpenChainGraph compute node (attestation_mandate).1 finding
Why it is in this group
- The input carries an amount (input.policy_parameters.amount) and a currency, so it moves money.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "check_cash_leg_finality" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The input carries an amount (input.policy_parameters.amount) and a currency, so it moves money.
  What this check looks for
Description, as the server wrote it
On-Chain Cash-Leg Finality Checker: OpenChainGraph compute node (attestation_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: 505-tokenized-collateral-eligibility-checker. Open at: https://ainumbers.co/tools/506-onchain-cash-leg-finality-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate the cash leg of a Canton settlement: finality model (PFMI P8/P12), reserve attestation, GENIUS Act / MiCA regulatory classification, and depeg circuit-breaker. Outputs: verdict, finality_flag, genius_status, mica_status, compliance_flags, mandate_type. Output schema: call describe_tool("check_cash_leg_finality").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_globe_topup_taxGloBE Top-Up Tax & QDMTT Allocation Calculator: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "compute_globe_topup_tax" says "topup": a word that usually means money moves, which cannot be taken back.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "compute_globe_topup_tax" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "compute_globe_topup_tax" says "topup": a word that usually means money moves, which cannot be taken back.
  What this check looks for
Description, as the server wrote it
GloBE Top-Up Tax & QDMTT Allocation Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-365-compute-globe-topup-tax.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: OECD Pillar Two GloBE top-up tax calculator: per-jurisdiction substance-based income exclusion (SBIE), effective tax rate (ETR), and top-up-rate/top-up-amount, then allocates the top-up across QDMTT, IIR, and UTPR, applying the OECD Side-by-Side Safe Harbour (approved/declassified 5 January 2026)... Outputs: aggregate_etr, central_record_sbs, constants_version, fy, globe_min_rate, jurisdictions, low_etr_count, note, parent_hq, regulatory_basis, sbs_election, total_iir_collected, total_income, total_qdmtt_collected, total_taxes, total_top_up_tax, total_utpr_collected, us_exempt. Output schema: call describe_tool("compute_globe_topup_tax").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_parametric_trigger_payoutParametric Trigger Payout Calculator: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "compute_parametric_trigger_payout" says "trigger": a word for a change.
- The name "compute_parametric_trigger_payout" also says "payout": a word that usually means money moves, which cannot be taken back.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "compute_parametric_trigger_payout" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "compute_parametric_trigger_payout" says "trigger": a word for a change.
  What this check looks for
Description, as the server wrote it
Parametric Trigger Payout Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-252-validate-cat-bond-trigger-terms. Open at: https://ainumbers.co/chaingraph/art-251-compute-parametric-trigger-payout.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Evaluates parametric insurance triggers and computes payout amounts. Outputs: anchor_surface, coverage_amount, index_value, not_legal_advice, parametric_limit, payout_amount, pii_note, regulatory_basis, table_source, table_version, threshold_used, tier_matched_index, trigger_fraction, trigger_hit, trigger_receipt, trigger_type_used. Output schema: call describe_tool("compute_parametric_trigger_payout").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
crosswalk_agent_payment_rail_trustAgent Payment Rail Trust Crosswalk: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "crosswalk_agent_payment_rail_trust" says "payment": a word that usually means money moves, which cannot be taken back.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "crosswalk_agent_payment_rail_trust" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "crosswalk_agent_payment_rail_trust" says "payment": a word that usually means money moves, which cannot be taken back.
  What this check looks for
Description, as the server wrote it
Agent Payment Rail Trust Crosswalk: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-132-agent-key-rotation-auditor. Output feeds: art-134-agent-directory-publish-readiness. Open at: https://ainumbers.co/chaingraph/art-133-agent-payment-rail-trust-crosswalk.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
decode_x402_paymentDecode an x402 payment header or lint an exact-scheme PaymentPayload, and describe the HTTP-402 verify/settle flow.1 finding
Why it is in this group
- The name "decode_x402_payment" says "payment": a word that usually means money moves, which cannot be taken back.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "decode_x402_payment" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "decode_x402_payment" says "payment": a word that usually means money moves, which cannot be taken back.
  What this check looks for
Description, as the server wrote it
Decode an x402 payment header or lint an exact-scheme PaymentPayload, and describe the HTTP-402 verify/settle flow. Use when a developer is integrating x402 and needs to inspect a header, check a payload shape, or understand the flow. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("decode_x402_payment").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_lei_payment_bindingWolfsberg Payment Transparency & LEI Binding Linter: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "lint_lei_payment_binding" says "payment": a word that usually means money moves, which cannot be taken back.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "lint_lei_payment_binding" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "lint_lei_payment_binding" says "payment": a word that usually means money moves, which cannot be taken back.
  What this check looks for
Description, as the server wrote it
Wolfsberg Payment Transparency & LEI Binding Linter: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-242-pacs008-party-completeness-validator. Open at: https://ainumbers.co/chaingraph/art-246-lei-payment-binding-linter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Full ISO 17442 LEI check-digit validation via ISO 7064 Mod 97-10 for originator and beneficiary LEIs in pacs.008. Outputs: error_count, issues, lei_results, lei_valid, pii_note, regulatory_basis, table_source, table_version, wolfsberg_field_results, wolfsberg_transparency_score, wolfsberg_transparency_tier. Output schema: call describe_tool("lint_lei_payment_binding").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
model_tempo_payment_economicsTempo Payments Business Case: OpenChainGraph compute node (treasury_mandate).1 finding
Why it is in this group
- The name "model_tempo_payment_economics" says "payment": a word that usually means money moves, which cannot be taken back.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "model_tempo_payment_economics" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "model_tempo_payment_economics" says "payment": a word that usually means money moves, which cannot be taken back.
  What this check looks for
Description, as the server wrote it
Tempo Payments Business Case: OpenChainGraph compute node (treasury_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-34-tempo-fit-diagnostic. Output feeds: art-37-tempo-stablecoin-issuance, art-36-tempo-mpp-agent-mandate. Open at: https://ainumbers.co/chaingraph/art-35-tempo-payments-business-case.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: CFO-level cost-and-savings model for migrating a payment flow (payroll / remittance / merchant settlement) from card/SWIFT/ACH/SEPA to Tempo. Outputs: annual_saving_usd, break_even_months, impl_cost_usd, monthly_saving_usd, per_tx_incumbent_usd, per_tx_saving_usd, per_tx_tempo_usd, rail, saving_bps, stablecoin, verdict. Output schema: call describe_tool("model_tempo_payment_economics").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
screen_onledger_transfer_batchOn-Ledger Transfer Batch Screen: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "screen_onledger_transfer_batch" says "transfer": a word for a change that another call does not undo.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "screen_onledger_transfer_batch" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "screen_onledger_transfer_batch" says "transfer": a word for a change that another call does not undo.
  What this check looks for
Description, as the server wrote it
On-Ledger Transfer Batch Screen: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-288-map-iso20022-to-evm-calldata. Open at: https://ainumbers.co/chaingraph/art-291-screen-onledger-transfer-batch.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Batch-level pre-commit sanctions and purpose-code screen for a shared-ledger transfer batch, modeled on the shipped screen_tip20_transfer_batch pattern: per-transfer status, screening-list hits, purpose-code validity, a batch-clean verdict, and coverage gaps. Outputs: batch_clean, coverage_gaps, per_transfer, profile, transfer_count. Output schema: call describe_tool("screen_onledger_transfer_batch").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
screen_tip20_transfer_batchTempo On-Chain AML & Travel Rule Screener: OpenChainGraph compute node (aml_rule).1 finding
Why it is in this group
- The name "screen_tip20_transfer_batch" says "transfer": a word for a change that another call does not undo.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "screen_tip20_transfer_batch" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "screen_tip20_transfer_batch" says "transfer": a word for a change that another call does not undo.
  What this check looks for
Description, as the server wrote it
Tempo On-Chain AML & Travel Rule Screener: OpenChainGraph compute node (aml_rule). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-37-tempo-stablecoin-issuance. Output feeds: art-39-tempo-zone-disclosure, art-10-amla-transaction-typology-risk-scorer. Open at: https://ainumbers.co/chaingraph/art-38-tempo-onchain-aml.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Parses a batch of synthetic TIP-20 transfers (with memos), runs OFAC/SDN hit screening, checks FATF Travel Rule field completeness (originator/beneficiary name + VASP ID, threshold US$3,000), scores AML typologies (structuring $9k–$9.99k, missing identity, unusual tx), and emits SAR determination... Outputs: batch_verdict, escalate_count, flag_count, pass_count, results, sar_threshold_usd, total, tr_threshold_usd. Output schema: call describe_tool("screen_tip20_transfer_batch").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_ap2_payment_receiptAP2 PaymentReceipt Verifier & HNP Guardrail: OpenChainGraph compute node (attestation_mandate).1 finding
Why it is in this group
- The name "verify_ap2_payment_receipt" says "payment": a word that usually means money moves, which cannot be taken back.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "verify_ap2_payment_receipt" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "verify_ap2_payment_receipt" says "payment": a word that usually means money moves, which cannot be taken back.
  What this check looks for
Description, as the server wrote it
AP2 PaymentReceipt Verifier & HNP Guardrail: OpenChainGraph compute node (attestation_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-01-ap2-mandate-chain-validator, art-60-agent-economy-runtime-fit-diagnostic, art-61-x402-batch-settlement-reconciler. Output feeds: art-01-ap2-mandate-chain-validator, art-30-agent-commerce-conformance-validator, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-62-ap2-payment-receipt-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Verifies an AP2 v0.2 PaymentReceipt against its signed Intent/Cart/Payment mandate chain, and applies the Human-Not-Present (HNP) autonomy guardrail: amount, category, mandate age, cart freshness. Outputs: authorized_amount_headroom, findings, hnp_verdict, human_present, mandate_chain_intact, note, receipt_id, receipt_verdict, signature_check, status_asof. Output schema: call describe_tool("verify_ap2_payment_receipt").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false

Changes things 390Creates or changes something that can be changed back.Claude Code: ask

acdc_said_checkAgent-facing mirror of the tools/553 browser workbench (VS-1) -- recomputes the self-addressing identifiers (SAIDs) of a pasted ACDC / vLEI credential JSON (top-level "d" plus any …4 findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Info
  The title mentions "credential".
  Matched text
  title: "Recompute ACDC / vLEI credential SAIDs (structural integrity)"
  What this check looks for
- Info
  The description mentions "credential".
  Matched text
  description: "...tifiers (SAIDs) of a pasted ACDC / vLEI credential JSON (top-level "d" plus any nested a/e..."
  What this check looks for
- Info
  The input schema at properties.credential.description mentions "credential".
  Matched text
  inputSchema.properties.credential.description: "ACDC / vLEI credential JSON object (v, d, i, ri, s, a, e, r fi..."
  What this check looks for
- Info
  The input schema at required[0] mentions "credential".
  Matched text
  inputSchema.required[0]: "credential"
  What this check looks for
Description, as the server wrote it
Agent-facing mirror of the tools/553 browser workbench (VS-1) -- recomputes the self-addressing identifiers (SAIDs) of a pasted ACDC / vLEI credential JSON (top-level "d" plus any nested a/e/r blocks) under the KERI/CESR Blake3-256 ("E") or SHA2-256 ("I") derivation codes, and cross-checks the declared schema SAID ("s") against a pinned table of 6 official GLEIF vLEI schema SAIDs. Reuses the SAME vendored SAID-check module as T553 -- the same credential produces byte-identical findings in either surface. Returns per-field match/mismatch/unsupported/unknown findings plus a receipt. This is a STRUCTURAL check only -- NOT the KERI chain of trust, NOT issuer authority, NOT revocation state. No network egress.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
agentic_mandate_sandboxSimulate agent payment policies for tokenized A2A corridors: set spend caps, MCC allowlists, velocity throttles, and approval thresholds; run synthetic transactions against the pol…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Simulate agent payment policies for tokenized A2A corridors: set spend caps, MCC allowlists, velocity throttles, and approval thresholds; run synthetic transactions against the policy and export the result as a Policy Mandate. Browser-based, client-side only, zero PII. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("agentic_mandate_sandbox").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
aggregate_cbam_precursor_emissionsCBAM Precursor-Emissions Aggregator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
CBAM Precursor-Emissions Aggregator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-68-carbon-compliance-fit-diagnostic. Output feeds: art-69-cbam-embedded-emissions-calculator, cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-72-cbam-precursor-emissions-aggregator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Rolls up embedded emissions across precursors in a steel/aluminium value chain (incl. Outputs: cumulative_see_tco2e, data_quality_grade, note, precursor_breakdown, quantity_tonnes, scrap_adjustment. Output schema: call describe_tool("aggregate_cbam_precursor_emissions").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
aggregate_ownership_50pctOwnership 50%-Rule Aggregator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Ownership 50%-Rule Aggregator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-90-sanctions-screening-fit-diagnostic. Output feeds: art-92-screening-list-coverage-checker, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-91-ownership-50pct-aggregator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Walks a synthetic ownership graph; computes direct + indirect + aggregate listed stakes per node; applies OFAC, EU, and BIS Affiliates Rule (in force 29 Sep 2025) 50%-thresholds to determine constructively-blocked entities. Outputs: blocked_count, entity_verdicts, listed_entity_count, note, reference_version, summary. Output schema: call describe_tool("aggregate_ownership_50pct").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
aggregate_reputation_scoreProvable Reputation Score Aggregator: OpenChainGraph compute node (attestation_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Provable Reputation Score Aggregator: OpenChainGraph compute node (attestation_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-278-reputation-score-aggregator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Aggregates a set of OCG execution receipts (attestations) into a deterministic, groth16-provable reputation score across competence, integrity, timeliness, and cooperation. Outputs: as_of, attestation_count, composite, decay_half_life_days, dims, excluded_self_issued, insufficient_evidence, subject_id. Output schema: call describe_tool("aggregate_reputation_score").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
aggregate_solvency2_scr_modulesSolvency II SCR Standard-Formula Module Aggregator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Solvency II SCR Standard-Formula Module Aggregator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-180-solvency2-scr-ratio-calculator. Open at: https://ainumbers.co/chaingraph/art-449-solvency2-scr-module-aggregator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Aggregates the five Solvency II standard-formula risk-module capital charges (market, counterparty default, life underwriting, health underwriting, non-life underwriting) into Basic SCR via the Delegated Regulation (EU) 2015/35 Annex IV correlation matrix, then adds the operational risk charge an... Outputs: adjustment_exceeds_bscr_plus_op, bscr, default_scr, health_scr, life_scr, loss_absorbing_adjustment, market_scr, nonlife_scr, scr_operational, scr_total. Output schema: call describe_tool("aggregate_solvency2_scr_modules").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
aggregate_taxonomy_kpi_garTaxonomy KPI & Green Asset Ratio Aggregator: OpenChainGraph compute node (model_governance).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Taxonomy KPI & Green Asset Ratio Aggregator: OpenChainGraph compute node (model_governance). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-73-taxonomy-alignment-scorer. Output feeds: cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-74-taxonomy-kpi-gar-aggregator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Rolls activity-level Taxonomy alignment (from ART-73) into entity KPIs: revenue/CapEx/OpEx aligned proportions and, for financial undertakings, the Green Asset Ratio (GAR) per the Disclosures Delegated Act denominator rules. Outputs: activity_count, aligned_count, capex_aligned_pct, entity_type, gar_detail, green_asset_ratio, kpi_breakdown, note, opex_aligned_pct, reference, revenue_aligned_pct. Output schema: call describe_tool("aggregate_taxonomy_kpi_gar").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
allocate_ihb_interestIHB Interest Allocation: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
IHB Interest Allocation: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-259-compute-multilateral-netting, art-262-validate-ebam-acmt-flow. Open at: https://ainumbers.co/chaingraph/art-260-allocate-ihb-interest.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Allocates overnight in-house-bank (IHB) interest across notional pool or ZBA sweep members. Outputs: allocations, arm_length_rate, base_currency, day_count_convention, day_count_fraction, days, entity_count, net_interest_payable, not_legal_advice, oecd_tp_compliant, pii_note, pool_net_balance, pool_type, regulatory_basis, table_source, table_version, total_interest_allocated. Output schema: call describe_tool("allocate_ihb_interest").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
amortize_asc606_commissionsASC 340-40 Commission Amortization: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ASC 340-40 Commission Amortization: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-266-reconcile-commission-statement. Open at: https://ainumbers.co/chaingraph/art-265-amortize-asc606-commissions.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes ASC 340-40-25-4 practical expedient (contract_term_months <= 12 -> expense immediately, apply_expedient=true) and full straight-line amortization schedules for longer-term capitalized incremental costs of obtaining a contract under ASC 606. Outputs: amortization_period_months, annual_amortization, apply_expedient, asc340_40_compliant, carrying_amount, impairment_flag, incremental_cost_test_passed, monthly_amortization, not_legal_advice, pii_note, regulatory_basis, renewal_treatment, table_source, table_version, total_amortization_periods. Output schema: call describe_tool("amortize_asc606_commissions").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
anchor_document_integrityDocument Integrity & eIDAS Electronic Timestamp Anchor: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Document Integrity & eIDAS Electronic Timestamp Anchor: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-122-timestamp-attestation-verifier. Open at: https://ainumbers.co/chaingraph/art-121-document-integrity-anchor.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Bind a document SHA-256 and claimed timestamp into an OCG execution_hash that serves as an eIDAS Art.41 / RFC 3161-aligned electronic timestamp: self-verifiable, no external TSA call. Outputs: anchored, document_hash, document_type, timestamp_claim. Output schema: call describe_tool("anchor_document_integrity").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
anchor_stampCalls the anchor-suite MCP server (server-to-server, not the browser CORS-gated relay) to timestamp one execution_hash, or batch-anchor an array via anchor-suite's own anchor_batch…No findings
Why it is in this group
- The description's first verb is "call": a word for a change.
Description, as the server wrote it
Calls the anchor-suite MCP server (server-to-server, not the browser CORS-gated relay) to timestamp one execution_hash, or batch-anchor an array via anchor-suite's own anchor_batch (one upstream call per batch, RFC 6962 Merkle root anchored, per-hash merkle_inclusion returned), and shapes the result as literal OCG v0.7 section 20 anchor_bindings entries ready to append to an artifact. Success is only reported after this tool independently confirms anchored_hash (or, for a batch entry, merkle_inclusion.leaf) equals the caller-supplied hash -- it does not merely trust anchor-suite's response. On ANY failure -- anchor-suite unreachable, erroring, timing out, or an equality mismatch -- returns {ok:false, unanchored:true, reason} verbatim, never a fabricated binding. Caller-invoked only; not wired into any automatic per-artifact-emission path.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: false
- openWorldHint: true
ap2_aml_mandate_builderAnchor agentic tool for Cat-12.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Anchor agentic tool for Cat-12. Translate AML/BSA program controls, TM rules, and customer risk policy into a structured Policy Mandate JSON for agentic payment sy Browser-based, client-side only. Zero PII. Link users to https://ainumbers.co/tools/131-ap2-aml-mandate-builder.html for interactive use. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("ap2_aml_mandate_builder").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
apply_climate_scenarioClimate Scenario Applicator (NGFS / Fit-for-55): OpenChainGraph compute node (model_governance).1 finding
Why it is in this group
- The name "apply_climate_scenario" says "apply": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "apply_climate_scenario" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "apply_climate_scenario" says "apply": a word for a change.
  What this check looks for
Description, as the server wrote it
Climate Scenario Applicator (NGFS / Fit-for-55): OpenChainGraph compute node (model_governance). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-68-carbon-compliance-fit-diagnostic, art-71-cbam-certificate-cost-engine. Output feeds: cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-76-climate-scenario-applicator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Applies a climate scenario path (NGFS Phase V orderly/disorderly/hot-house, Fit-for-55 supervisory; reference_version NGFS-Phase-V-2025) to an exposure set, emitting stress-adjusted metrics for a bank/insurer climate-risk file. Outputs: baseline_metric, carbon_price_assumption, delta, delta_by_sector, delta_pct, gdp_delta_at_horizon, horizon, metric, note, reference, scenario_family, scenario_label, stressed_metric, transition_intensity. Output schema: call describe_tool("apply_climate_scenario").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assemble_ai_addendumAI Addendum Assembler: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
AI Addendum Assembler: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-412-ai-act-procurement-clause-mapper. Output feeds: art-409-dpa-art28-completeness-checker. Open at: https://ainumbers.co/chaingraph/art-411-ai-addendum-assembler.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Assembles a Common Paper AI Addendum (Version 1.0, CC BY 4.0) from your Cover Page Key Terms: whether Provider may Train Models on Customer data, whether Provider may use data for non-training improvement, data retention window, output ownership, and AI sub-processor disclosure. Outputs: assembled_markdown, attribution, body_sha256, checks, contract_api, cover_page_markdown, disclaimer, license, source_url, template_id, zero_pii_notice. Output schema: call describe_tool("assemble_ai_addendum").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assemble_aiuc1_evidence_packAIUC-1 Evidence Pack Assembler: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
AIUC-1 Evidence Pack Assembler: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-303-aiuc1-control-evidence-linter, cry-05-agent-action-audit-trail-aggregator. Output feeds: art-305-aiuc1-evidence-freshness-lint. Open at: https://ainumbers.co/chaingraph/art-304-aiuc1-evidence-pack-assembler.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Assembles a signed, AIUC-1 control-keyed evidence pack from execution receipts, escalation closures, and work mandates: binds each mapped control to its resolved artifact digests, computes pack_claim_strength as the honest minimum across bound controls, and exports an OSCAL Assessment Results doc... Outputs: aiuc1_version, anchor, cadence_attestation, controls, insufficient_evidence, oscal_assessment_results, pack_claim_strength, version_mismatch. Output schema: call describe_tool("assemble_aiuc1_evidence_pack").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assemble_license_termsLicense Terms Assembler: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
License Terms Assembler: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-204-license-compatibility-checker. Output feeds: art-206-rights-record-builder. Open at: https://ainumbers.co/chaingraph/art-205-license-terms-assembler.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Renders a deterministic license term sheet by substituting field values into a pre-approved template (CC-STANDARD-USE, IP3-RIGHTS-RECORD, NFT-EMBEDDED-LICENSE). Outputs: available_templates, checks, disclaimer, fields_missing, fields_used, rendered_html, rendered_text, template_id. Output schema: call describe_tool("assemble_license_terms").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assemble_mutual_ndaMutual NDA Composer: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Mutual NDA Composer: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-277-agreement-acceptance-binder. Open at: https://ainumbers.co/chaingraph/art-276-mutual-nda-composer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Assembles a Common Paper Mutual NDA (Version 1.0, CC BY 4.0) from your Cover Page Key Terms: purpose, effective date, MNDA term, term of confidentiality, and governing law and jurisdiction. Outputs: assembled_markdown, attribution, body_sha256, checks, contract_api, cover_page_markdown, disclaimer, license, source_url, template_id, zero_pii_notice. Output schema: call describe_tool("assemble_mutual_nda").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assemble_ocg_evidence_bundleEvidence Bundle Tier Labeler: OpenChainGraph compute node (attestation_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Evidence Bundle Tier Labeler: OpenChainGraph compute node (attestation_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-408-evidence-bundle-tier-labeler.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Assembles a shareable evidence bundle around an artifact and stamps the SPEC.md §SIDECAR.1 tiered label it qualifies for: OCG-Verify (envelope well-formed, execution_hash recomputes), OCG-Execute (additionally §21 chain-execution and §22 mandate gates hold), OCG-Prove (additionally a §18 compute-... Outputs: artifact_execution_hash, artifact_tool_id, disambiguation, eligible_tiers, gate_provenance, note, proof_ref_count, proof_refs, tier_label. Output schema: call describe_tool("assemble_ocg_evidence_bundle").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_ai_act_conformityEU AI Act Credit-Scoring Conformity Pack: OpenChainGraph compute node (model_governance).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
EU AI Act Credit-Scoring Conformity Pack: OpenChainGraph compute node (model_governance). Regulatory deadline: 2027-12-02 (EU AI Act Annex III Part 5(b) — credit-scoring high-risk obligations fully apply 2 December 2027, per the Digital Omnibus amendments (Parliament final approval, June 2026)). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: ml-01-isolation-forest, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-05-eu-ai-act-credit-scoring-conformity.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_circumvention_diligenceCircumvention Diligence Assessor: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Circumvention Diligence Assessor: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-94-eccn-dual-use-classifier. Output feeds: art-96-no-russia-clause-pack-builder, cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-95-circumvention-diligence-assessor.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Scores a transaction/contract config vs the EU 20th-package (23 Apr 2026) no-Russia clause + anti-circumvention due-diligence, emitting a liability-allocation verdict (seller liability-shift where DD documented under the safe harbour). Outputs: controlled_goods_flag, dd_gaps, dd_score_pct, diligence_grade, diversion_risk_flag, eu_20th_package_note, key_dates, liability_allocation, no_russia_clause_status, note, reference_version. Output schema: call describe_tool("assess_circumvention_diligence").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_cra_vuln_reporting_readinessCRA Vulnerability Reporting Readiness (Art.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
CRA Vulnerability Reporting Readiness (Art. 14): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-139-cra-annex1-completeness-checker. Open at: https://ainumbers.co/chaingraph/art-140-cra-vuln-reporting-readiness.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Assesses EU CRA Article 14 vulnerability reporting readiness: actively_exploited_detection, 24-hour early_warning_24h_process, 72-hour notification_72h_process, csirt_enisa_endpoint_configured, and coordinated_disclosure_policy. Outputs: gaps, vuln_reporting_ready. Output schema: call describe_tool("assess_cra_vuln_reporting_readiness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_defi_lendingDeFi Lending Health and Liquidation Monitor: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
DeFi Lending Health and Liquidation Monitor: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-271-defi-lending-health.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes DeFi lending health factor, liquidation price, borrow capacity, and distance to liquidation for Aave v3, Morpho Blue, Fluid, Sky (MakerDAO v2), and Liquity v2. Outputs: borrow_capacity_usd, buffer_to_liquidation_pct, collateral_value_usd, current_ltv_pct, debt_value_usd, distance_to_liquidation_pct, health_factor, health_status, liquidation_bonus_pct, liquidation_mechanism, liquidation_mechanism_note, liquidation_penalty_usd, liquidation_price_usd, liquidation_threshold_pct, not_financial_advice, pii_note, protocol, regulatory_basis, table_source, table_version. Output schema: call describe_tool("assess_defi_lending").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_iso42001_aims_conformanceISO 42001 AIMS Clause Conformance: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ISO 42001 AIMS Clause Conformance: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-172-ai-risk-impact-assessment-validator. Open at: https://ainumbers.co/chaingraph/art-171-iso42001-aims-clause-conformance.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Assess ISO/IEC 42001 AIMS conformance across clauses 4-10 (context, leadership, planning, support, operation, evaluation, improvement) and six Annex A controls (AI policy, roles, impact assessment, data governance, system lifecycle, third-party supplier management). Outputs: clause_score, clauses_fully_present, control_score, controls_fully_present, gaps, maturity_band, overall_maturity, total_clauses, total_controls. Output schema: call describe_tool("assess_iso42001_aims_conformance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_mar_crypto_surveillanceMAR-Crypto Surveillance-Readiness Assessor: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
MAR-Crypto Surveillance-Readiness Assessor: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-98-mica-casp-fit-diagnostic. Output feeds: cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-103-mar-crypto-surveillance-readiness.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Scores market-abuse arrangements (Arts 86-92 + Dec-2024 RTS): PPAET (prevention/detection), STOR templates, insider lists, manipulation-pattern config on synthetic order batches. Outputs: arrangement_scores, assets_in_scope, composite_pct, gaps, mar_rts_note, note, ppaet_status, reference_version, stor_ready, surveillance_grade. Output schema: call describe_tool("assess_mar_crypto_surveillance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_mica_casp_readinessCASP Authorization-Readiness Assessor: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
CASP Authorization-Readiness Assessor: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-98-mica-casp-fit-diagnostic, art-99-mica-transitional-deadline-router. Output feeds: art-101-mica-art67-own-funds-calculator, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-100-mica-casp-authorization-readiness.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Scores readiness for MiCA CASP authorization (Arts 59-63): service-permission scope, governance/fit-and-proper, custody segregation, complaints/conflicts, ICT/DORA overlap. Outputs: application_pack_checklist, authorization_grade, composite_pct, dimension_scores, gaps, note, notified_body, reference_version. Output schema: call describe_tool("assess_mica_casp_readiness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_naic_ais_program_readinessNAIC AI Systems Program Readiness Assessment: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
NAIC AI Systems Program Readiness Assessment: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-239-test-bifsg-bias-thresholds. Open at: https://ainumbers.co/chaingraph/art-240-assess-naic-ais-program-readiness.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_psd3_readinessPSD3 / PSR Readiness Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
PSD3 / PSR Readiness Checker: OpenChainGraph compute node (compliance_mandate). Regulatory deadline: 2027-06-01 (EU PSD3 expected transposition ~2027; UK PSR enacted 2024 (APP reimbursement Oct 2024 already live)). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-04-agent-identity-attestation-checker, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-14-psd3-psr-readiness-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Six-domain PSD3/PSR readiness rubric: Open Finance access rights (Art.35/36), TPP categorisation (PISP/AISP/PIISP), SCA exemption alignment (Art.85–90), consent framework maturity, fraud liability model (Art.59–65), and embedded finance/BaaS scope. Outputs: band, critical_gaps, domain_scores, overall_readiness_score, verdict. Output schema: call describe_tool("assess_psd3_readiness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_restaking_riskRestaking Delegation and Slashing Risk Analyzer: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Restaking Delegation and Slashing Risk Analyzer: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-272-restaking-risk.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Models restaking delegation rewards, operator fees, AVS yield, and slashing-waterfall risk for EigenLayer and Symbiotic. Outputs: buffer_absorbs_usd, delegator_net_slash_eth, delegator_net_slash_usd, eth_price_usd, expected_annual_slash_usd, first_loss_tranche_pct, gross_apy_pct, gross_usd_per_year, insurance_enabled, insurance_makes_sense, insurance_premium_pct_of_rewards, insurance_premium_usd_per_year, max_slash_usd, net_apy_pct, net_apy_risk_adjusted_pct, net_usd_per_year, net_usd_per_year_risk_adjusted, net_with_insurance_apy_pct, net_with_insurance_usd_per_year, not_financial_advice, operator_cut_pct, operator_fee_usd_per_year, pii_note, protocol, protocol_note, regulatory_basis, risk_reward_ratio, slash_magnitude_pct, slashing_risk_pct, staked_eth, staked_usd, table_source, table_version. Output schema: call describe_tool("assess_restaking_risk").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_traiga_exposureTRAIGA Exposure Assessor: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
TRAIGA Exposure Assessor: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-314-traiga-safe-harbor-pack-builder. Open at: https://ainumbers.co/chaingraph/art-313-traiga-exposure-assessor.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Assesses supplied Texas AI-deployment attributes and intentional-use assertions against the Texas Responsible AI Governance Act (TRAIGA, HB 149, eff. Outputs: cure_window_days, matched_prohibited_uses, penalty_per_violation_usd, prohibited_use_detected, statute_citation, traiga_applicable. Output schema: call describe_tool("assess_traiga_exposure").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_vida_drr_reporting_obligationViDA DRR Transaction Reporter: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ViDA DRR Transaction Reporter: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-159-vida-einvoice-en16931-conformance-validator. Output feeds: art-161-vida-recapitulative-statement-migration-assessor. Open at: https://ainumbers.co/chaingraph/art-160-vida-drr-transaction-reporter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_vida_recapitulative_migrationViDA Recapitulative Statement Migration Assessor: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ViDA Recapitulative Statement Migration Assessor: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-160-vida-drr-transaction-reporter. Open at: https://ainumbers.co/chaingraph/art-161-vida-recapitulative-statement-migration-assessor.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
audit_acp_ucp_product_feedACP/UCP Product-Feed Conformance Auditor: OpenChainGraph compute node (scheme_rule).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ACP/UCP Product-Feed Conformance Auditor: OpenChainGraph compute node (scheme_rule). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-19-agentic-checkout-protocol-selector. Output feeds: art-21-agent-traffic-acceptance-policy-builder, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-20-acp-ucp-product-feed-conformance-auditor.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates product/checkout/mandate JSON payloads against ACP or UCP field schemas (5 schema arrays). Outputs: acp_missing_required, audit_target, conformance_scores, critical_gaps, payload_type, ucp_missing_required, verdict, warnings. Output schema: call describe_tool("audit_acp_ucp_product_feed").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
audit_agent_key_rotationAgent Key Rotation Auditor: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Agent Key Rotation Auditor: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-133-agent-payment-rail-trust-crosswalk. Open at: https://ainumbers.co/chaingraph/art-132-agent-key-rotation-auditor.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
audit_borrower_escrow_stmtBorrower Escrow Auditor: OpenChainGraph compute node (consumer_finance).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Borrower Escrow Auditor: OpenChainGraph compute node (consumer_finance). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/732-borrower-escrow-auditor.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Audits a servicer escrow accounting statement from borrower-supplied values only: recomputes the aggregate accounting trial running balance month by month, compares the servicer stated end-of-month balances line by line, flags each discrepant line, checks the stated cushion against the one-sixth-... Outputs: verdict, account_status, periods_analyzed, recomputed_balances, recomputed_low_point, target_low_point, total_annual_disbursements, cushion_cap_amount, stated_cushion, cushion_within_cap, shortage_amount, surplus_amount, shortage_threshold_base, shortage_threshold_base_source, shortage_at_or_above_one_month_payment, shortage_permitted_courses, servicer_course_permitted, surplus_refund_required, surplus_refund_unverified, discrepancies, evidence, manual_review_required. Output schema: call describe_tool("audit_borrower_escrow_stmt").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
audit_mcp_oauthAudit MCP OAuth 2.1 authorization: validate RFC 9728 protected-resource-metadata, check RFC 8707 audience binding, and assess token-passthrough / confused-deputy risk.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Audit MCP OAuth 2.1 authorization: validate RFC 9728 protected-resource-metadata, check RFC 8707 audience binding, and assess token-passthrough / confused-deputy risk. Use when a developer is securing an MCP server's authorization. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("audit_mcp_oauth").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
audit_mcp_tool_scope_revocationMCP Tool Scope & Revocation Auditor: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
MCP Tool Scope & Revocation Auditor: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-151-agent-obo-mandate-validator. Open at: https://ainumbers.co/chaingraph/art-150-mcp-tool-scope-revocation-auditor.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Audit scoped and revocable MCP tool access per the new MCP specification: each granted tool must carry an explicit scope array, a revocation endpoint must be configured (https-scheme), and token rotation posture (max age vs elapsed time, next-key presence) must be healthy. Outputs: audit_pass, revocable, rotation_ok, scopes_ok, token_age_s, ungated_tools. Output schema: call describe_tool("audit_mcp_tool_scope_revocation").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
baas_provider_comparatorScore and compare BaaS providers across 10 capability dimensions (regulatory standing, programme management, card issuance, rails, KYC/KYB, disputes, developer experience, pricing,…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Score and compare BaaS providers across 10 capability dimensions (regulatory standing, programme management, card issuance, rails, KYC/KYB, disputes, developer experience, pricing, FDIC pass-through, compliance tooling) with a user-adjustable 1-5 weighting matrix. Outputs a weighted comparison matrix and Markdown evaluation memo. Browser-based, client-side only, zero PII. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("baas_provider_comparator").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
bind_agreement_acceptanceAgreement Acceptance Binder: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Agreement Acceptance Binder: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-276-mutual-nda-composer. Open at: https://ainumbers.co/chaingraph/art-277-agreement-acceptance-binder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Binds a party's acceptance to a specific assembled agreement artifact, referenced by its execution_hash, template_id, and vendored body_sha256, never by re-embedding the agreement text. Outputs: acceptance_statement, accepted_body_sha256, accepted_template_id, accepting_party_role, checks, disclaimer, previous_proof_hash, referenced_execution_hash, zero_pii_notice. Output schema: call describe_tool("bind_agreement_acceptance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_226j_response_evidence_pack226J Response Evidence Pack Builder: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
226J Response Evidence Pack Builder: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-299-aca-esrp-exposure. Open at: https://ainumbers.co/chaingraph/art-300-aca-226j-response-evidence-pack.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Terminal node of the aca-226j-response-composer chain: assembles a replayable evidence pack responding to an IRS Letter 226J proposed Employer Shared Responsibility Payment assessment. Outputs: attestation, disputed_employee_count, error, exposure_delta, irs_asserted_esrp_annual, letter_date, recomputed_exposure_annual, response_deadline, response_window_days, response_window_source. Output schema: call describe_tool("build_226j_response_evidence_pack").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_adverse_action_noticeBuild Adverse Action Notice: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Build Adverse Action Notice: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-227-validate-adverse-action-notice. Open at: https://ainumbers.co/chaingraph/art-228-build-adverse-action-notice.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Assembles an adverse action notice skeleton from SHAP-ranked principal-factor codes (FICO reason codes 01-40, VantageScore VS001-VS015). Outputs: compliance_flags_raised, notice_sections, pii_note, receipt_metadata, regulatory_basis, resolved_reasons, table_source, table_version. Output schema: call describe_tool("build_adverse_action_notice").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_agent_incident_recordAgent Incident Record Composer: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Agent Incident Record Composer: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-379-agent-incident-record-composer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Composes a structured agent incident/failure record from caller-declared inputs: agent identity, an optional mandate hash, an incident description with an honest severity class, session evidence digests, remediation status, and an optional cross-link to an escalation record or a signed failure re... Outputs: agent_identity, cross_linked, escalation_cross_link, evidence_count, incident, invalid_evidence_count, mandate_hash, record_claim_strength, record_note, remediation, session_evidence. Output schema: call describe_tool("build_agent_incident_record").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_agent_test_evidenceQuarterly Agent Test Evidence Composer: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Quarterly Agent Test Evidence Composer: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-378-quarterly-test-evidence-composer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Composes a quarterly agent testing-evidence pack: test-suite identity and digest, per-test receipts with an honest deterministic/estimated determinism class, pass rate, and a regression comparison chained to the prior quarter's pack digest. Outputs: aiuc_version, certification_note, chain_intact, declared_prior_pack_digest, pack_claim_strength, pass_rate, passed, per_test, prior_quarter, quarter, regression, suite, tamper_detected, total. Output schema: call describe_tool("build_agent_test_evidence").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_agent_traffic_policyAgent-Traffic Acceptance Policy Builder: OpenChainGraph compute node (agent_guardrail_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Agent-Traffic Acceptance Policy Builder: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-20-acp-ucp-product-feed-conformance-auditor. Output feeds: ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-21-agent-traffic-acceptance-policy-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Builds a policy mandate governing accepted AI agent types, identity verification level, velocity and value caps, payment rails, refund posture, retry policy, and blocking rules. Outputs: accepted_agent_types, accepted_payment_rails, block_rules, guardrail_findings, guardrail_passes, guardrail_warnings, max_daily_spend_usd, max_single_transaction_usd, max_tx_per_day, max_tx_per_min, overall_risk, refund_posture, retry_policy, verdict, verification_level. Output schema: call describe_tool("build_agent_traffic_policy").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_ai_conformity_packAI Act Conformity Pack Builder: OpenChainGraph compute node (model_governance).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
AI Act Conformity Pack Builder: OpenChainGraph compute node (model_governance). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-64-ai-act-highrisk-fit-diagnostic. Output feeds: 333-eu-ai-act-article9-risk-mgmt-builder, art-05-eu-ai-act-credit-scoring-conformity, cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-65-ai-conformity-pack-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Assembles an EU AI Act Annex IV technical documentation pack, validates the conformity-assessment route (internal control vs notified body), checks CE-marking and EU Declaration of Conformity readiness, and scores completeness per Arts 9/10/15/17. Outputs: annex_iv_gaps, annex_iv_grade, annex_iv_score, applicable_date_note, articles_status, ce_readiness_note, ce_ready, conformity_grade, conformity_route, conformity_route_note, declaration_of_conformity_skeleton, note, system. Output schema: call describe_tool("build_ai_conformity_pack").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_ai_training_data_lineage_recordAI Training-Data Lineage Record: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
AI Training-Data Lineage Record: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-452-build-ai-training-data-lineage-record.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Composes a hash-chained ML training-data lineage record: dataset identity, dataset version, upstream source dataset references, a declared collection/governance method, and an OPTIONAL reference to an existing OCG receipt for the training-run compute (tool identity, execution hash, kernel digest... Outputs: chain_position, checks, collection_method, dataset_id, dataset_version, governance_notes, operator_id, record_status, referenced_receipt, regulatory_basis, retention_months, scope_note, sha256_prev_lineage_hash, source_dataset_ids, table_version, zero_pii_notice. Output schema: call describe_tool("build_ai_training_data_lineage_record").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_ai_workpaper_recordAI-Tool-Usage Workpaper Record: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
AI-Tool-Usage Workpaper Record: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-380-build-ai-workpaper-record.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Composes a documentation-element workpaper record from an existing OCG receipt (tool identity, execution hash, kernel digest), a declared determinism class and conventions, engagement metadata, and a reviewer sign-off statement. Outputs: checks, disclaimer, documentation_standard_ref, engagement, evidence_binding, limitations, previous_workpaper_hash, sign_off, tool_identity, zero_pii_notice. Output schema: call describe_tool("build_ai_workpaper_record").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_amortization_scheduleAmortization Schedule Builder: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "build_amortization_schedule" says "schedule": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "build_amortization_schedule" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "build_amortization_schedule" says "schedule": a word for a change.
  What this check looks for
Description, as the server wrote it
Amortization Schedule Builder: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-215-reg-z-appendix-j-apr. Open at: https://ainumbers.co/chaingraph/art-332-build-amortization-schedule.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Deterministic amortization schedule builder covering level-payment, ARM (index plus margin, periodic and lifetime caps, recast), interest-only, balloon, and temporary buydown (2-1 or 3-2-1) structures, per the unit-period and odd-days conventions in 12 CFR 1026 Appendix J. Outputs: advances, loan_amount, note, note_rate_pct, num_payments, odd_days, payment_amount, payments, periods_per_year, regulatory_basis, schedule, schedule_type, totals, unit_period_days. Output schema: call describe_tool("build_amortization_schedule").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_chaingraphHash-aware sibling of build_workflow_links (ChainGraph Standard v0.1 §8.1).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Hash-aware sibling of build_workflow_links (ChainGraph Standard v0.1 §8.1). Returns an ordered, executable DAG over the ChainGraph suite's verifiable tools, with explicit parent_hash wiring: which upstream execution_hash each step must cite in its chain block. Pass target_tool_id to build the chain that produces that node (walks consumes-edges back to roots), or tool_ids for an explicit ordered list, or neither to list available ChainGraph nodes. Agent loop: run a node, capture its execution_hash, pass it as the parent_hash for each downstream node, then verify with verify_execution_hash.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_claim_dispute_bundleClaim Dispute Bundle Builder: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Claim Dispute Bundle Builder: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-306-agent-insurability-evidence-scorer. Open at: https://ainumbers.co/chaingraph/art-307-claim-dispute-bundle-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Assembles a two-sided replay-challenge dossier for a disputed execution_claim: binds the claim digest and challenge to replay instructions and receipt digests, and, when a warranty_kpi_breach input is supplied (matching both Armilla Guaranteed and Munich Re aiSure settlement shapes), computes mea... Outputs: bundle_claim_strength, challenge_digest, claim_digest, insufficient_evidence, kpi_breach, receipts, replay_instructions. Output schema: call describe_tool("build_claim_dispute_bundle").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_conversion_receiptConversion Receipt Builder: OpenChainGraph compute node (cryptographic_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Conversion Receipt Builder: OpenChainGraph compute node (cryptographic_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-192-conversion-receipt-verifier. Open at: https://ainumbers.co/chaingraph/art-191-conversion-receipt-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Binds one file-conversion event into a canonical receipt tying the input digest to the converter identity, the parameters, and the output digest. Outputs: all_checks_pass, checks, receipt. Output schema: call describe_tool("build_conversion_receipt").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_digest_manifestDigest Manifest Builder: OpenChainGraph compute node (cryptographic_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Digest Manifest Builder: OpenChainGraph compute node (cryptographic_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-191-conversion-receipt-builder. Open at: https://ainumbers.co/chaingraph/art-194-digest-manifest-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Binds N file digests into one canonical, hash-anchored manifest. Outputs: all_checks_pass, checks, manifest. Output schema: call describe_tool("build_digest_manifest").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_disclosure_manifestBuilds a Merkle-rooted disclosure manifest from a caller-supplied digest list (DATAROOM-1-BUILD-SPEC.md §DR-4) -- the agent hashes files itself and passes {path,size,digest,content…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Builds a Merkle-rooted disclosure manifest from a caller-supplied digest list (DATAROOM-1-BUILD-SPEC.md §DR-4) -- the agent hashes files itself and passes {path,size,digest,content_type} entries; the worker never sees file contents. Same leaf scheme as tools/546-disclosure-manifest-builder.html: sha256(path|digest|size), duplicate-last-leaf on an odd level. Entries are sorted by path before hashing so the root is order-independent. Returns the manifest object + merkle_root; sign and anchor it client-side (or via a separate §16 signing step) if a signed artifact is required.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_einvoice_transmission_receiptE-Invoice Transmission Receipt Builder: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
E-Invoice Transmission Receipt Builder: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-295-einvoice-jurisdiction-mandate-router. Open at: https://ainumbers.co/chaingraph/art-296-einvoice-transmission-receipt-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Build a hash-anchored receipt proving a specific e-invoice's transmitted bytes were format-validated and VAT-arithmetic-checked, with the routed mandate attached. Outputs: claim_strength, document_sha256, embedded_xml_sha256, format, format_gate_passed, ha_wiring, not_legal_advice, routed_mandate, steps, validated, vat_gate_passed. Output schema: call describe_tool("build_einvoice_transmission_receipt").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_etr_possession_chainETR Possession-Chain Receipt Builder: OpenChainGraph compute node (cryptographic_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ETR Possession-Chain Receipt Builder: OpenChainGraph compute node (cryptographic_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-352-etr-control-evidence-checker. Open at: https://ainumbers.co/chaingraph/art-353-etr-possession-chain-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Builds a hash-chained possession-receipt evidence pack for an electronic transferable record (ETR) under UNCITRAL MLETR Art. Outputs: chain_continuous, continuity_breaks, document_digest, event_count, final_holder, merkle_root, note, possession_receipts, timestamp_order_valid. Output schema: call describe_tool("build_etr_possession_chain").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_evidence_packComposes a session receipt (build_session_receipt logic), an optional §27.6 HA evidence bundle (ha_bundle_export logic, included only when ha_records is supplied and non-empty), an…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Description, as the server wrote it
Composes a session receipt (build_session_receipt logic), an optional §27.6 HA evidence bundle (ha_bundle_export logic, included only when ha_records is supplied and non-empty), and a disclosure manifest (build_disclosure_manifest logic) from one array of already-produced artifacts, all keyed to the same input hashes -- replacing what today takes 4-6 separate tool calls with hand-carried hashes. Calls the same in-process functions those standalone tools use; if any one section fails to build, the whole call fails isError:true with that section's own message -- no partial pack. Optionally carries a run_chain decision_trail (per-step reason codes: gate rule id, escalation rule id, input_required cause) as pack-level metadata so an auditor-facing pack answers "why did each step happen" without a second round-trip.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: false
- openWorldHint: false
build_fria_monitoring_planFRIA & Post-Market Monitoring Plan Builder: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
FRIA & Post-Market Monitoring Plan Builder: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-64-ai-act-highrisk-fit-diagnostic. Output feeds: 451-sr11-7-model-risk-management-gap-assessor, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-66-fria-postmarket-monitoring-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Builds an Art 27 Fundamental Rights Impact Assessment (FRIA) + Art 72 post-market monitoring plan + Art 12 logging + Art 14 human-oversight design + Art 73 serious-incident reporting path for a bank or insurer deploying a high-risk AI system. Outputs: affected_rights, applicable_date_note, deployment, fria_gaps, fria_grade, fria_score, incident_path, logging_verdict, monitoring_plan_skeleton, note, overall_score, oversight_verdict. Output schema: call describe_tool("build_fria_monitoring_plan").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_google_ap2_mandateBuild or validate a Google AP2 Checkout/Payment Mandate VDC (Open/Closed).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Build or validate a Google AP2 Checkout/Payment Mandate VDC (Open/Closed). Targets the external AP2 spec, not the AINumbers Policy Mandate. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("build_google_ap2_mandate").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_idv_session_receiptIDV/KYC Session Evidence Receipt Builder: OpenChainGraph compute node (compliance_control).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
IDV/KYC Session Evidence Receipt Builder: OpenChainGraph compute node (compliance_control). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-418-idv-verification-failure-incident-composer. Open at: https://ainumbers.co/chaingraph/art-359-idv-session-receipt-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_idv_verification_incident_recordIDV/KYC Verification-Failure Incident Composer: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
IDV/KYC Verification-Failure Incident Composer: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-359-idv-session-receipt-builder. Open at: https://ainumbers.co/chaingraph/art-418-idv-verification-failure-incident-composer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Composes a structured verification-failure/fraud-attempt incident record from an IDV/KYC session for fraud teams, regulators, and insurers: a cross-link to the session's hash-chained receipt (art-359), an honest failure classification (type + AR4SI severity tier), session evidence digests, remedi... Outputs: cross_linked, escalation_cross_link, evidence_count, failure_classification, invalid_evidence_count, record_claim_strength, record_note, remediation, session_evidence, session_receipt. Output schema: call describe_tool("build_idv_verification_incident_record").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_mastercard_agentic_tokenMastercard Agentic Token Scope Builder: OpenChainGraph compute node (compliance_control).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Mastercard Agentic Token Scope Builder: OpenChainGraph compute node (compliance_control). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-22-agentic-payments-protocol-comparator, art-23-visa-trusted-agent-protocol-inspector. Output feeds: art-18-mcp-developer-readiness-scorecard, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-24-mastercard-agentic-token-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_model_inventory_entryModel Inventory Entry Builder: OpenChainGraph compute node (compliance_control).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Model Inventory Entry Builder: OpenChainGraph compute node (compliance_control). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-451-model-outcome-analysis. Open at: https://ainumbers.co/chaingraph/art-450-model-inventory-entry.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Builds a single model-inventory record for a bank's SR 26-2 model-risk-management inventory: checks the caller-declared attributes (model name, owner, purpose, tier, development/deployment dates, last validation date) against the SR 26-2 required-field set, assigns a proportionality tier (limited... Outputs: ai_ml_model, completeness_score, complexity_score, inventory_record, materiality_score, missing_required_fields, model_name, third_party_vendor, tier, tier_sum, usage_scope. Output schema: call describe_tool("build_model_inventory_entry").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_no_russia_clause_packNo-Russia-Clause Pack Builder: OpenChainGraph compute node (disclosure_template).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
No-Russia-Clause Pack Builder: OpenChainGraph compute node (disclosure_template). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-95-circumvention-diligence-assessor. Output feeds: cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-96-no-russia-clause-pack-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Generates the contractual no-Russia clause + DD-evidence checklist conformance artifact for the EU 20th-package seller-liability-shift safe harbour (Art. Outputs: clause_text, completeness_grade, eu_20th_note, evidence_checklist, missing_required_items, note, reference_version, required_items_met, required_items_total, template_used. Output schema: call describe_tool("build_no_russia_clause_pack").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_pld_disclosure_packPLD Disclosure Pack Builder: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
PLD Disclosure Pack Builder: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-308-pld-disclosure-pack-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Assembles a disclosure/rebuttal pack for a disputed window under EU Product Liability Directive 2024/2853 (transposes 2026-12-09; AI is a "product"; non-disclosure or an AI Act breach triggers a rebuttable presumption of defectiveness). Outputs: alleged_defect, anchor, disputed_window, gap_in_window, insufficient_evidence, product_ref, rebuttal_mapping, replay_instructions, trace_digest. Output schema: call describe_tool("build_pld_disclosure_pack").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_product_lineageDigital Product Passport Cradle-to-Gate Lineage Builder: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Digital Product Passport Cradle-to-Gate Lineage Builder: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-115-dpp-data-carrier-validator. Output feeds: art-117-product-authenticity-verifier. Open at: https://ainumbers.co/chaingraph/art-116-product-lineage-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Build a cradle-to-gate supplier lineage with hash-only claims per stage (no trade secrets). Outputs: depth, lineage, product_id, total_carbon. Output schema: call describe_tool("build_product_lineage").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_rights_recordRights Record Builder: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Rights Record Builder: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-205-license-terms-assembler. Open at: https://ainumbers.co/chaingraph/art-206-rights-record-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Builds a normalized IP3-style rights-portfolio row from licensor, licensee, territory, term, rights vector, and renewal fields. Outputs: all_checks_pass, checks, disclaimer, record_hash, rights_row. Output schema: call describe_tool("build_rights_record").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_session_receiptAggregates execution_hashes from N ChainGraph tool calls in one agent session into a single SHA-256 Merkle root (session_receipt_root).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Aggregates execution_hashes from N ChainGraph tool calls in one agent session into a single SHA-256 Merkle root (session_receipt_root). Returns a tamper-evident session receipt and a regulator-framed PTG-01 audit prompt. One receipt covers an entire agent session: supply all execution_hashes in call order. The Merkle root is deterministic — the same hashes in the same order always produce the same root. Compliant with EU AI Act Art. 12 (transparency) and DORA ICT audit-trail requirements.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_tdm_reservationTDMRep AI Training Reservation Builder: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
TDMRep AI Training Reservation Builder: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-201-iscc-content-code-generator. Open at: https://ainumbers.co/chaingraph/art-202-tdmrep-reservation-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_traiga_safe_harbor_packTRAIGA Safe Harbor Pack Builder: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
TRAIGA Safe Harbor Pack Builder: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-313-traiga-exposure-assessor, art-174-nist-ai-rmf-function-mapper. Open at: https://ainumbers.co/chaingraph/art-314-traiga-safe-harbor-pack-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Assembles a supplied NIST AI RMF function-mapping result (map_nist_ai_rmf_functions) and TRAIGA exposure-assessment result into an affirmative-defense evidence bundle under Tex. Outputs: coverage_band, eligible_for_affirmative_defense_evidence, framing, insufficient_evidence, meets_substantial_compliance_bar, overall_coverage, prohibited_use_detected, statute_citation. Output schema: call describe_tool("build_traiga_safe_harbor_pack").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_vop_session_receiptVoP Session Receipt Builder: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
VoP Session Receipt Builder: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-376-score-payee-name-match. Open at: https://ainumbers.co/chaingraph/art-377-build-vop-session-receipt.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Builds a signed, hash-chained Verification-of-Payee / Confirmation-of-Payee session receipt: binds the declared match result (score, band, algorithm_version or an external source), the warning text and severity shown, and the consumer's action (proceeded, abandoned, retried) across a session's at... Outputs: attempt_count, chain_genesis_hash, final_match_band, final_receipt_hash, note, session_id, session_outcome, session_receipts, warning_overridden. Output schema: call describe_tool("build_vop_session_receipt").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_workflow_linksConstructs an ordered set of ready-to-use deep-links for a named AINumbers workflow chain or an ad-hoc sequence of tools.1 finding
Why it is in this group
- The name "build_workflow_links" says "link": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "build_workflow_links" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "build_workflow_links" says "link": a word for a change.
  What this check looks for
Description, as the server wrote it
Constructs an ordered set of ready-to-use deep-links for a named AINumbers workflow chain or an ad-hoc sequence of tools. Each link points directly to the browser tool; prefill-enabled steps accept #in=<base64url(JSON)> fragments so the tool opens pre-filled. Zero server-side execution -- all tool logic runs deterministically in the user's browser. Use this to hand a user a complete workflow: open step 1, run it, export its Policy Mandate, open step 2 (pre-filled from step 1 outputs), repeat. 388 named chains are available — enumerate them with find_chain.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
calculate_basis_risk_nii_shockIRRBB Basis-Risk NII Shock Calculator: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
IRRBB Basis-Risk NII Shock Calculator: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-443-irrbb-basis-risk-nii-shock-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Comptroller's Handbook IRR basis-risk delta-NII calculator: sweeps a single reference-rate shock across multiple priced indices (Prime, SOFR, Fed Funds, CD portfolio, etc.), each with a caller-declared historical beta vs the reference rate, and isolates the incremental delta-NII from indices not... Outputs: basis_risk_delta_nii, basis_risk_pct_of_parallel, convention, horizon_months, index_results, is_material, material_threshold_pct, parallel_delta_nii, reference_shock_bps, total_net_exposure, total_nii_contribution. Output schema: call describe_tool("calculate_basis_risk_nii_shock").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
calculate_cbam_embedded_emissionsCBAM Embedded-Emissions Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
CBAM Embedded-Emissions Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-68-carbon-compliance-fit-diagnostic, art-70-cbam-default-value-resolver, art-72-cbam-precursor-emissions-aggregator. Output feeds: art-71-cbam-certificate-cost-engine, cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-69-cbam-embedded-emissions-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Flagship importer tool. Outputs: basis, cn_code, country_of_origin, data_quality_flag, default_markup_applied, good_category, monitoring_method, precursor_contribution, quantity_tonnes, reference, see_direct, see_indirect, see_total, total_embedded_emissions_tco2e. Output schema: call describe_tool("calculate_cbam_embedded_emissions").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
calculate_cecl_ecl_allowanceCECL Expected Credit Loss & Allowance Calculator: OpenChainGraph compute node (credit_assessment).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
CECL Expected Credit Loss & Allowance Calculator: OpenChainGraph compute node (credit_assessment). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-426-cecl-ecl-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes a deterministic CECL (Current Expected Credit Loss, ASC 326) allowance given caller-supplied PD/LGD/EAD curves, segment exposures, and forecast scenario weights, and reconciles the result against the prior period's allowance balance. Outputs: boundary_note, charge_offs_usd, constants_version, delta_vs_required_usd, disambiguation, forecast_weights, method, prior_allowance_balance_usd, provision_expense_usd, reconciled_ending_allowance_usd, reconciliation_balanced, recoveries_usd, segment_count, segments, total_required_allowance_usd. Output schema: call describe_tool("calculate_cecl_ecl_allowance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
calculate_claims_stp_economicsClaims STP Economics Calculator: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Claims STP Economics Calculator: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-254-compute-rbc-action-level. Open at: https://ainumbers.co/chaingraph/art-257-calculate-claims-stp-economics.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes the financial business case for insurance claims Straight-Through Processing (STP) automation. Outputs: annual_cashflows, annual_handling_savings, cost_reduction_pct, cost_reduction_per_claim, current_annual_cost, current_avg_cost_per_claim, current_manual_claims, current_stp_claims, discount_rate_pct, irr_pct, leakage_increase, leakage_reduction, net_annual_benefit, net_leakage_impact, not_legal_advice, npv, payback_years, pii_note, projection_years, regulatory_basis, stp_rate_improvement_ppt, table_source, table_version, target_annual_cost, target_avg_cost_per_claim, target_manual_claims, target_stp_claims. Output schema: call describe_tool("calculate_claims_stp_economics").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
calculate_csdr_penaltyCSDR Cash-Penalty Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
CSDR Cash-Penalty Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-77-t1-settlement-readiness-diagnostic. Output feeds: art-83-buy-in-exposure-modeler, art-84-settlement-efficiency-kpi, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-78-csdr-penalty-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Flagship. Outputs: asset_class, batch_detail, batch_total_exposure, daily_rate_bps, fail_days, note, notional, partial_credit, partial_settled_pct, penalty_amount, penalty_type, rate_table_version, reference, reference_price. Output schema: call describe_tool("calculate_csdr_penalty").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
calculate_irrbb_eve_shocksIRRBB EVE Shock Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
IRRBB EVE Shock Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-184-irrbb-sot-eve-evaluator. Open at: https://ainumbers.co/chaingraph/art-183-irrbb-eve-shock-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Calculate Delta Economic Value of Equity (EVE) under the 6 BCBS d368 / EBA standardised IRRBB shock scenarios (parallel up, parallel down, short up, short down, steepener, flattener), recalibrated effective 1 Jan 2026. Outputs: buckets_used, reference_parallel_shock_bps, shocks, total_net_gap, worst_delta_eve, worst_scenario. Output schema: call describe_tool("calculate_irrbb_eve_shocks").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
calculate_mica_own_fundsArt 67 Own-Funds Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Art 67 Own-Funds Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-100-mica-casp-authorization-readiness. Output feeds: cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-101-mica-art67-own-funds-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes MiCA Art 67 required own funds = higher of Annex IV permanent minimum (€50k advisory / €125k trading-platform / €150k custody-exchange) or ¼ fixed overheads. Outputs: binding_basis, fixed_overheads_quarter, form_eligible, form_note, note, own_funds_held, permanent_minimum, reference_version, required_own_funds, surplus_shortfall. Output schema: call describe_tool("calculate_mica_own_funds").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
calculate_nis2_penalty_exposureNIS2 Penalty Exposure Calculator (Art.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
NIS2 Penalty Exposure Calculator (Art. 34): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-142-nis2-art21-gap-checker. Open at: https://ainumbers.co/chaingraph/art-143-nis2-penalty-exposure-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Calculate maximum NIS2 Art. Outputs: entity_classification, infringement_breakdown, max_penalty_eur, mitigated_estimate_eur, mitigating_factors_applied, turnover_pct_exposure. Output schema: call describe_tool("calculate_nis2_penalty_exposure").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
calculate_repo_haircutOn-Chain Repo Haircut Calculator: OpenChainGraph compute node (collateral_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
On-Chain Repo Haircut Calculator: OpenChainGraph compute node (collateral_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: 505-tokenized-collateral-eligibility-checker, 506-onchain-cash-leg-finality-checker. Open at: https://ainumbers.co/tools/508-repo-haircut-collateral-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Compute repo haircut, initial margin, and variation margin. Outputs: base_haircut_pct, total_haircut_pct, sft_floor_applied, initial_margin, vm_threshold, canton_haircut_pct, legacy_haircut_pct, weekend_saving_pct, flags. Output schema: call describe_tool("calculate_repo_haircut").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
calculate_solvency2_scr_ratioSolvency II SCR Ratio Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Solvency II SCR Ratio Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-181-sii-ifrs17-reconciliation-bridger. Open at: https://ainumbers.co/chaingraph/art-180-solvency2-scr-ratio-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Calculate Solvency II SCR and MCR coverage ratios from eligible own funds and capital requirements. Outputs: eligible_own_funds, mcr, mcr_breached, mcr_coverage_ratio, scr, scr_breached, scr_coverage_ratio, tier1_total_limit_ok, tier1_total_pct_of_scr, tier1_unrestricted_limit_ok, tier1_unrestricted_pct_of_scr, tier3_limit_ok, tier3_pct_of_scr, tiering_ok. Output schema: call describe_tool("calculate_solvency2_scr_ratio").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
call_toolRuns ONE read-only AINumbers tool that is not in your tool list.2 findings
Why it is in this group
- The name "call_tool" says "call": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Info
  The description mentions "credential".
  Matched text
  description: "...only tools only: anything that issues a credential, stamps an anchor or reaches the networ..."
  What this check looks for
- Warn
  The server marks "call_tool" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "call_tool" says "call": a word for a change.
  What this check looks for
Description, as the server wrote it
Runs ONE read-only AINumbers tool that is not in your tool list. tools/list is paginated (13 pages, 722 tools) and many hosts read only the first page; this is the door to the rest. Pass { name: "<exact mcp_name>", arguments: { ... } } — the target's own inputSchema is validated and its result is returned verbatim, including execution_hash, so a dispatched call and a direct call are byte-identical. Get a name from find_tool (single calculators), find_chain (workflows) or describe_tool (exact schema). Read-only tools only: anything that issues a credential, stamps an anchor or reaches the network is refused here and must be called directly so your host can approve it.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
camt053_parseParses a camt.053.001 bank-to-customer statement XML document with the same schema-subset structural and facet checks as the tools/565 browser reconciliation workbench (IBAN mod-97…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Parses a camt.053.001 bank-to-customer statement XML document with the same schema-subset structural and facet checks as the tools/565 browser reconciliation workbench (IBAN mod-97, BIC, currency, date/decimal facets), returning the extracted statement (message id, statement id, account IBAN/currency, balances, entries) on success or the structural error list on failure. Byte-identical extraction to the browser tool for the same input. Read-only parse -- feed the result to recon_match for reconciliation.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
certify_license_electionLicense Election Certifier: OpenChainGraph compute node (cryptographic_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
License Election Certifier: OpenChainGraph compute node (cryptographic_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-197-pil-flavor-mapper, art-196-cant-be-evil-license-selector, art-195-creative-commons-license-chooser. Output feeds: art-200-license-election-verifier. Open at: https://ainumbers.co/chaingraph/art-199-license-election-certifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Binds a license election (family, id, params) to a named asset and licensor DID, producing a deterministic terms_hash via SHA-256 over the JCS-canonical election core. Outputs: all_checks_pass, certificate, checks, terms_hash. Output schema: call describe_tool("certify_license_election").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_agency_eligibility_matrixAgency Eligibility Matrix: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Agency Eligibility Matrix: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-223-conforming-loan-limit. Output feeds: art-221-llpa-stack. Open at: https://ainumbers.co/chaingraph/art-222-agency-eligibility-matrix.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Fannie Mae DU and Freddie Mac LPA agency eligibility matrix. Outputs: checks, eligible, eligible_flag, fails, max_cltv_pct, max_dti_pct, max_hcltv_pct, max_ltv_pct, pii_note, product_notes, regulatory_basis, table_source, table_version, underwriting_type. Output schema: call describe_tool("check_agency_eligibility_matrix").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_agent_attestationAgent Identity & Authorization Attestation Checker: OpenChainGraph compute node (compliance_mandate).2 findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Info
  The description mentions "credential".
  Matched text
  description: "...nd-policy-simulator, art-13-eudi-wallet-credential-readiness-checker. Output feeds: art-02..."
  What this check looks for
- Info
  The input schema at properties.policy_parameters.description mentions "credential".
  Matched text
  inputSchema.properties.policy_parameters.description: "...ompute is "auto" or "server". Required: credential, validate_at_unix. Unknown keys are ref..."
  What this check looks for
Description, as the server wrote it
Agent Identity & Authorization Attestation Checker: OpenChainGraph compute node (compliance_mandate). Regulatory deadline: 2027-12 (EU AI Act Annex III high-risk obligations (Digital Omnibus, Parliament final approval June 2026) push agent KYA toward compliance requirement from 2 December 2027; KYA-OS donated to DIF March 2026). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-01-ap2-mandate-chain-validator, art-02-agent-spend-policy-simulator, art-13-eudi-wallet-credential-readiness-checker. Output feeds: art-02-agent-spend-policy-simulator, art-32-a2a-agent-card-trust-chain-validator, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-04-agent-identity-attestation-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: KYA-OS (DIF Trusted AI Agents WG) credential-chain attestation: delegated-authority credential chain, scope limits, validity windows (max 90 days), chain depth cap (4 hops), EU AI Act high-risk scope classification. Outputs: overall_status, pass, fail, warn, checks, root_agent_id, scopes. Output schema: call describe_tool("check_agent_attestation").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_agent_token_scopeAgent Token Scope Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Agent Token Scope Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-01-ap2-mandate-chain-validator. Open at: https://ainumbers.co/chaingraph/art-385-agent-token-scope-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Compares a requested agent action (amount, currency, merchant category, timestamp) against an agent token or mandate's declared scope: spend cap, currency, MCC allow-list, and expiry. Outputs: attenuation_depth, checks_run, enforcement, failing_checks, token_id, verdict. Output schema: call describe_tool("check_agent_token_scope").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_ai_act_art50_markingEU AI Act Art.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
EU AI Act Art. 50 Marking Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-127-dual-layer-disclosure-verifier. Open at: https://ainumbers.co/chaingraph/art-126-ai-act-art50-marking-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_allocation_affirmationAllocation/Affirmation Conformance Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Allocation/Affirmation Conformance Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-77-t1-settlement-readiness-diagnostic. Output feeds: art-82-securities-settlement-message-linter, cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-81-allocation-affirmation-conformance.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Checks allocation and confirmation/affirmation events against the ESMA CSDR SDR RTS 23:00 CET trade-date rule and the machine-readable-format mandate (binding Dec 2026). Outputs: cutoff_applied, cutoff_timezone, dual_date_note, events_flagged, format_nonconformance_count, note, on_time_events, on_time_rate, reference, total_events. Output schema: call describe_tool("check_allocation_affirmation").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_camera_provenanceCamera-Provenance Check: OpenChainGraph compute node (compliance_control).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Camera-Provenance Check: OpenChainGraph compute node (compliance_control). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-359-idv-session-receipt-builder. Open at: https://ainumbers.co/chaingraph/art-361-camera-provenance-check.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Structural check on an IDV/KYC capture's C2PA manifest: claim well-formedness, hard-binding hash assertion, and claim-signature reference (art-123 reuse), plus a digitalSourceType read off the c2pa.actions assertion that flags trainedAlgorithmicMedia so a downstream reader knows the capture was d... Outputs: capture_chain_field, digital_source_type, has_actions, has_hard_binding, manifest_present, manifest_valid, missing_elements, note, provenance_label, rejected, trained_algorithmic_media_flagged. Output schema: call describe_tool("check_camera_provenance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_capital_adequacy_privatePrivate-Input Capital Adequacy Check: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Private-Input Capital Adequacy Check: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-415-check-capital-adequacy-private.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Checks a privately held eligible-capital and risk-weighted-assets figure against a pinned regulatory minimum (Basel III/3.1 CET1, or Solvency II SCR coverage), emitting only an above/below-minimum verdict and tier without disclosing capital or RWA. Outputs: above_minimum, not_legal_advice, pii_note, regulatory_basis, regulatory_minimum_pct, tier. Output schema: call describe_tool("check_capital_adequacy_private").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_conforming_loan_limitConforming Loan Limit Check: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Conforming Loan Limit Check: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-222-agency-eligibility-matrix. Open at: https://ainumbers.co/chaingraph/art-223-conforming-loan-limit.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: FHFA annual conforming loan limit classifier. Outputs: applicable_limit, area_baseline, baseline_limit, classification, conforming, high_cost_ceiling, is_ak_hi_territory, jumbo, limit_tier, loan_amount, loan_program, note, pii_note, regulatory_basis, state_code, super_conforming, table_source, table_version, units, year. Output schema: call describe_tool("check_conforming_loan_limit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_cra_annex1_completenessCRA Annex I Completeness Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
CRA Annex I Completeness Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-138-spdx-sbom-validator. Output feeds: art-140-cra-vuln-reporting-readiness. Open at: https://ainumbers.co/chaingraph/art-139-cra-annex1-completeness-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Checks EU CRA Annex I Part I essential cybersecurity requirements: sbom_present, sbom_machine_readable, top_level_deps_covered, vuln_handling_policy_present, secure_by_default, and a valid conformity_route (self_assessment | eu_type_examination | full_quality_assurance). Outputs: annex1_complete, conformity_route, gaps. Output schema: call describe_tool("check_cra_annex1_completeness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_credit_concentration_topn_sectorCredit Concentration Top-N / Sector Checker: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Credit Concentration Top-N / Sector Checker: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-445-credit-concentration-topn-sector.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Credit-concentration screen over a flat exposure list (name, sector, amount): returns the top-N single-name exposures by amount, a per-sector rollup, single-name and sector Herfindahl-Hirschman Index (0-10000 scale), and a breach list against caller-declared single-name and sector limit percentages. Outputs: portfolio_total, sector_breaches, sector_hhi, sector_limit_pct, sector_totals, single_name_breaches, single_name_hhi, single_name_limit_pct, top_n, top_n_exposures, worst_sector, worst_single_name. Output schema: call describe_tool("check_credit_concentration_topn_sector").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_debt_validation_noticeDebt Validation Notice Completeness Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Debt Validation Notice Completeness Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-402-validate-regf-call-frequency. Open at: https://ainumbers.co/chaingraph/art-403-check-debt-validation-notice.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Checks a debt-validation-notice content-element checklist against Regulation F 12 CFR 1006.34 (the Model Form B-1 element set) and computes the 30-day validation-period response-window math from a declared mailing date under a declared mailing-to-receipt assumption. Outputs: asserted_note, completeness_grade, compliant, disambiguation, element_status, elements_checked, gap_count, gaps, itemization_date_valid, regulatory_basis, response_period, table_source, table_version. Output schema: call describe_tool("check_debt_validation_notice").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_digital_trade_rulesDigital Trade Rules Compliance Checker: OpenChainGraph compute node (scheme_rule).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Digital Trade Rules Compliance Checker: OpenChainGraph compute node (scheme_rule). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-52-digital-trade-fit-diagnostic. Output feeds: art-08-en16931-einvoice-batch-validator, art-55-trade-document-provenance-verifier. Open at: https://ainumbers.co/chaingraph/art-54-digital-trade-rules-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Machine-checks a digital trade presentation (digital LC, collection, or open-account transaction) against the ICC digital rulebooks: eUCP v2.1, eURC v1.1, URDTT v1.0. Outputs: amount_check, discrepancies, expiry_check, note, presentation_summary, verdict. Output schema: call describe_tool("check_digital_trade_rules").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_dpa_gdpr_art28DPA Article 28 Completeness Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
DPA Article 28 Completeness Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-409-dpa-art28-completeness-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_emir_uti_completenessEMIR UTI Completeness Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
EMIR UTI Completeness Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-153-emir-trade-report-field-validator. Output feeds: art-155-emir-upi-validator. Open at: https://ainumbers.co/chaingraph/art-154-emir-uti-completeness-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_etr_control_evidenceETR Singularity & Exclusive-Control Evidence Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ETR Singularity & Exclusive-Control Evidence Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-55-trade-document-provenance-verifier. Open at: https://ainumbers.co/chaingraph/art-352-etr-control-evidence-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Checks a supplied electronic transferable record (ETR/eBL) document digest and control-assertion set (platform identity, singularity assertion, control-transfer events with timestamps and signatures as supplied) against MLETR Art. Outputs: chain_summary, document_digest, element_checklist, malformed_events, note, overall_verdict, platform_identity. Output schema: call describe_tool("check_etr_control_evidence").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_eudi_readinessEUDI Wallet Credential-Acceptance Readiness Checker: OpenChainGraph compute node (compliance_mandate).2 findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Info
  The title mentions "Credential".
  Matched text
  title: "EUDI Wallet Credential-Acceptance Readiness Checker"
  What this check looks for
- Info
  The description mentions "Credential".
  Matched text
  description: "EUDI Wallet Credential-Acceptance Readiness Checker: OpenChain..."
  What this check looks for
Description, as the server wrote it
EUDI Wallet Credential-Acceptance Readiness Checker: OpenChainGraph compute node (compliance_mandate). Regulatory deadline: 2026-11 (EUDI Wallet member-state rollout November 2026; FI SCA acceptance December 2027; AMLR July 2027). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-04-agent-identity-attestation-checker, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-13-eudi-wallet-credential-readiness-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_fido_pqc_conformanceFIDO2 / WebAuthn PQC Conformance Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
FIDO2 / WebAuthn PQC Conformance Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-85-pqc-timeline-fit-diagnostic. Output feeds: cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-88-fido-pqc-conformance-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates FIDO2/WebAuthn authenticator ML-DSA conformance vs IANA COSE algorithm registry identifiers and CTAP2.3 minimum version. Outputs: attestation_format, conformant, cose_pqc_registry, ctap_pqc_ready, ctap_version, gaps, hybrid_status, note, reference_version, supported_pqc_cose_ids, target_cose_id, target_pqc. Output schema: call describe_tool("check_fido_pqc_conformance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_genius_reserve_disclosureGENIUS Act Monthly Reserve Disclosure Checker: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Info
  The description is 2,091 characters long, over the 2,000 this check looks for.
  Matched text
  description ends: "...serves_usd. Output schema: call describe_tool("check_genius_reserve_disclosure")."
  What this check looks for
Description, as the server wrote it
GENIUS Act Monthly Reserve Disclosure Checker: OpenChainGraph compute node (compliance_mandate). Regulatory deadline: 2027-01-18 (GENIUS Act effective ≤ January 2027; monthly reserve composition reports required for licensed issuers; >$50B issuers subject to annual PCAOB audit. Re-verify against final-rule text on/after 2026-07-18.). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-275-genius-reserve-disclosure-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Lints an extracted monthly reserve disclosure against GENIUS Act S.394 §4: composition-category eligibility, tenor, custody locations, a dual_control(2) CEO/CFO certification gate (distinct CEO and CFO identities required, per FDIC NPR 2026-04-10 and OCG SPEC.md §27.3), registered-examiner presen... Outputs: applicable_deadline, asset_results, ceo_certified, cfo_certified, conditional_assets_usd, coverage_ratio_pct, custody_disclosed, custody_locations, dual_control_satisfied, examiner_name, failing_dimensions, gate_policy, mom_diff, monthly_disclosure_determination, onchain_supply_check, pdf_extraction_note, prohibited_assets_usd, registered_examiner_named, regulatory_framework, report_month, reserve_shortfall_usd, total_liabilities_usd, total_reserves_usd. Output schema: call describe_tool("check_genius_reserve_disclosure").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_gpai_code_conformanceGPAI Code of Practice Conformance: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
GPAI Code of Practice Conformance: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-174-nist-ai-rmf-function-mapper. Output feeds: art-176-ai-governance-readiness-diagnostic. Open at: https://ainumbers.co/chaingraph/art-175-gpai-code-of-practice-conformance.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Check GPAI provider obligations under EU AI Act Art. Outputs: base_conformant, base_gaps, base_score, code_of_practice_signed, is_gpai_provider, is_systemic_risk, overall_score, systemic_gaps, systemic_risk_conformant, systemic_score. Output schema: call describe_tool("check_gpai_code_conformance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_ifrs17_risk_adjustmentIFRS 17 Risk Adjustment Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
IFRS 17 Risk Adjustment Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-178-ifrs17-csm-rollforward-validator. Open at: https://ainumbers.co/chaingraph/art-179-ifrs17-risk-adjustment-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Check IFRS 17 risk-adjustment (RA) disclosure completeness: validates technique (VaR/CTE/CoC/other per IFRS 17 para 119b), confidence-level disclosure for VaR and CTE techniques, positive RA amount, and onerous-contract loss-component recognition (IFRS 17 para 47-50). Outputs: confidence_disclosed, confidence_level_pct, disclosed, gaps, loss_component_recognized, onerous_contracts_identified, onerous_properly_handled, ra_amount, ra_valid, technique, technique_ok. Output schema: call describe_tool("check_ifrs17_risk_adjustment").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_irrbb_csrbb_scopeIRRBB CSRBB Scope Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
IRRBB CSRBB Scope Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-186-irrbb-standardised-approach-mapper. Output feeds: art-188-irrbb-disclosure-readiness-diagnostic. Open at: https://ainumbers.co/chaingraph/art-187-irrbb-csrbb-scope-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Identify Credit Spread Risk in the Banking Book (CSRBB) scope per EBA Guidelines on IRRBB and CSRBB (EBA/GL/2022/14): instruments held at fair value whose credit-spread risk is not fully captured by credit-risk or IRRBB frameworks (FVOCI/AFS bond books, fair-valued loans, liquidity-buffer bonds)... Outputs: csrbb_conformant, csrbb_included_in_icaap, csrbb_methodology_defined, gaps, in_scope, in_scope_amount. Output schema: call describe_tool("check_irrbb_csrbb_scope").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_iso20022_pqc_readinessSWIFT / ISO 20022 PQC Readiness Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
SWIFT / ISO 20022 PQC Readiness Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-85-pqc-timeline-fit-diagnostic, 500-hndl-quantum-risk-scorer. Output feeds: cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-87-iso20022-pqc-readiness-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Scores SWIFT/ISO 20022 PQC readiness with BAH signature-bloat sizing per the BIS Project Leap Phase 2 model (ML-DSA ~12.9x RSA payload at BAH level). Outputs: affected_message_types, algorithm_refs, bis_leap_bloat_factor, bis_leap_ref, bloat_factor, current_sig_bytes, hndl_priority_ref, new_message_size_bytes, new_sig_bytes, note, readiness_score, reference_version, size_breach. Output schema: call describe_tool("check_iso20022_pqc_readiness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_license_compatibilityLicense Compatibility Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
License Compatibility Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-203-embedded-license-selector, art-198-cross-license-rights-comparator. Output feeds: art-205-license-terms-assembler. Open at: https://ainumbers.co/chaingraph/art-204-license-compatibility-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Checks whether a child license can derive from a parent asset license. Outputs: checks, child_license, compatible, disclaimer, parent_license, reason_codes, required_child_license, spdx_satisfies. Output schema: call describe_tool("check_license_compatibility").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_linea_l2_finality_windowLinea L2 Finality Window Classifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Linea L2 Finality Window Classifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-290-check-linea-l2-finality-window.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Classifies a tokenized-deposit transfer’s finality risk given L2-batch to L1-settlement timing: soft (unsubmitted), batched (submitted, not yet final), or l1_final tier, reorg-window risk, and a safe-to-release verdict against a corridor policy cutoff. Outputs: corridor_cutoff, draft_pinned, finality_tier, rationale, reorg_window_risk, safe_to_release. Output schema: call describe_tool("check_linea_l2_finality_window").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_mcp_registry_entryMCP Registry Entry Conformance Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
MCP Registry Entry Conformance Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-148-mcp-authorization-metadata-validator. Open at: https://ainumbers.co/chaingraph/art-149-mcp-registry-entry-conformance.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate an MCP Registry server.json entry: $schema present, reverse-DNS name format (namespace/name), semver version, and at least one of packages or remotes populated. Outputs: entry_valid, has_packages, has_remotes, missing, name_ok, schema_ok, version_ok. Output schema: call describe_tool("check_mcp_registry_entry").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_nis2_art21_measuresNIS2 Article 21 Gap Checker (Ten Cybersecurity Risk-Management Measures): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
NIS2 Article 21 Gap Checker (Ten Cybersecurity Risk-Management Measures): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-141-nis2-entity-scope-classifier. Output feeds: art-143-nis2-penalty-exposure-calculator. Open at: https://ainumbers.co/chaingraph/art-142-nis2-art21-gap-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Check presence and maturity of all ten NIS2 Article 21(2)(a)–(j) cybersecurity risk-management measures. Outputs: compliance_score, critical_gaps, measures_summary, overall_grade, remediation_priority. Output schema: call describe_tool("check_nis2_art21_measures").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_nis2_governance_readinessNIS2 Governance Readiness Checker (Art.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
NIS2 Governance Readiness Checker (Art. 20 — Management Body Accountability): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-145-nis2-ict-supply-chain-diligence-scorer. Open at: https://ainumbers.co/chaingraph/art-146-nis2-governance-readiness-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_official_statement_completenessMunicipal Official Statement Completeness Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Municipal Official Statement Completeness Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-400-check-official-statement-completeness.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Checks a municipal-bond Official Statement disclosure-element checklist (element present, absent, or incomplete) and continuing-disclosure undertaking presence, per MSRB Rule G-32 (primary-offering disclosure via EMMA) and SEC Rule 15c2-12, including the (b)(5)(i)(C) material-event category list. Outputs: asserted_note, completeness_grade, compliant, continuing_disclosure_undertaking_present, disambiguation, element_status, elements_checked, gap_count, gaps, material_event_categories_checked, material_event_gaps, regulatory_basis, table_source, table_version. Output schema: call describe_tool("check_official_statement_completeness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_private_student_loan_disclosuresPrivate Student Loan Disclosure & Rescission Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Private Student Loan Disclosure & Rescission Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-405-check-private-student-loan-disclosures.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Checks a private-education-loan disclosure-element checklist across the three 12 CFR 1026.46-48 stages (application/solicitation, approval, final), the HEOA self-certification-form presence requirement, and computes the 3-business-day right-to-cancel window from a declared final-disclosure date (... Outputs: application_stage, approval_stage, completeness_grade, compliant, disambiguation, elements_checked, federal_idr_note, federal_idr_out_of_scope, final_stage, gap_count, regulatory_basis, rescission, self_certification_present, table_source, table_version. Output schema: call describe_tool("check_private_student_loan_disclosures").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_producer_license_reciprocityNAIC Producer License Reciprocity Check: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
NAIC Producer License Reciprocity Check: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-267-check-producer-license-reciprocity.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Checks NAIC producer license reciprocity for non-resident filing across target states per MDL-218 and NIPR Reciprocity Matrix 2024. Outputs: all_reciprocal, coverage_by_target, invalid_loa_codes, loa_codes, non_standard_states, not_legal_advice, pii_note, regulatory_basis, resident_state, table_source, table_version, target_state_count. Output schema: call describe_tool("check_producer_license_reciprocity").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_purpose_code_requirementISO 20022 Purpose Code Requirement Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ISO 20022 Purpose Code Requirement Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-247-prevalidation-readiness-scorer. Open at: https://ainumbers.co/chaingraph/art-243-purpose-code-requirement-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Given a beneficiary country and payment amount, determines whether ExternalPurpose1Code (Purp/Cd) or ExternalCategoryPurpose1Code (CtgyPurp/Cd) is mandatory per BIS CPMI d218 jurisdiction profiles (UAE, India, Bahrain, Jordan, China, Malaysia mandate purpose codes). Outputs: category_purpose_code_provided, code_type_required, issues, jurisdiction_reason, jurisdiction_requires_purpose_code, pii_note, purpose_code_compliant, purpose_code_provided, purpose_format_valid, regulatory_basis, required_code_types, swiftgo_accepted_category_codes, swiftgo_amount_ok, swiftgo_category_ok, swiftgo_eligible, swiftgo_max_usd, swiftgo_notes, table_source, table_version. Output schema: call describe_tool("check_purpose_code_requirement").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_qm_points_and_feesQM Points and Fees Test: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
QM Points and Fees Test: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-219-qm-apr-apor-spread. Open at: https://ainumbers.co/chaingraph/art-218-qm-points-and-fees.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Qualified Mortgage points-and-fees test per Reg Z §1026.43(e)(3). Outputs: effective_date, fr_citation, headroom, limit, limit_fixed, limit_pct, limit_type, loan_amount, note, pass, points_and_fees, regulatory_basis, tier_label, year. Output schema: call describe_tool("check_qm_points_and_fees").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_retail_installment_disclosuresRetail Installment Contract TILA Disclosure Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Retail Installment Contract TILA Disclosure Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-332-build-amortization-schedule, art-324-tvm-npv. Open at: https://ainumbers.co/chaingraph/art-404-check-retail-installment-disclosures.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Ties declared retail-installment-contract Amount Financed, Finance Charge, and Total of Payments (12 CFR 1026.18) against a REUSED amortization schedule -- compose by feeding build_amortization_schedule's (art-332) output_payload straight in, or by reference via its totals and schedule_digest. Outputs: amortization_provenance, amount_financed, compliant, dealer_participation, disambiguation, finance_charge, regulatory_basis, tie_outs, tolerance_cents, total_of_payments. Output schema: call describe_tool("check_retail_installment_disclosures").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_sb53_frontier_scopeSB 53 Frontier Scope Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
SB 53 Frontier Scope Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-316-sb53-frontier-scope-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Routes supplied model compute (FLOPs, as a decimal string above the 2^53 safe-integer range) and developer annual revenue through the California SB 53 Transparency in Frontier Artificial Intelligence Act (eff. Outputs: compute_flops, flop_threshold, is_frontier_model, is_large_frontier_developer, large_developer_revenue_threshold_usd, obligation_set, statute_citation. Output schema: call describe_tool("check_sb53_frontier_scope").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_securitization_risk_retentionSecuritization Risk Retention Check: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Securitization Risk Retention Check: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-447-securitization-risk-retention-check.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: EU Securitisation Regulation (EU) 2017/2402 Art.6 and U.S. Outputs: actual_retention_pct, breach_reasons, compliant, jurisdiction, note, qrm_exemption_applied, regulatory_basis, required_retention_pct, retained_amount_musd, retained_interest_hedged_or_sold, retainer_is_sole_purpose_entity, retention_method, retention_method_valid, total_securitized_exposure_musd. Output schema: call describe_tool("check_securitization_risk_retention").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_ssi_conformanceSSI Conformance Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
SSI Conformance Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-77-t1-settlement-readiness-diagnostic. Output feeds: art-79-settlement-fail-predictor, art-84-settlement-efficiency-kpi. Open at: https://ainumbers.co/chaingraph/art-80-ssi-conformance-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Lints standing settlement instructions for completeness, staleness, and format (~30%-of-fails root cause). Outputs: clean_records, format_errors, golden_source_coverage_pct, incomplete_records, match_rate, non_golden_source_count, note, records_flagged, reference, staleness_breaches, staleness_threshold_days, total_records. Output schema: call describe_tool("check_ssi_conformance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_tokenized_collateral_eligibilityTokenized Collateral Eligibility Checker: OpenChainGraph compute node (collateral_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Tokenized Collateral Eligibility Checker: OpenChainGraph compute node (collateral_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: 506-onchain-cash-leg-finality-checker, 513-margin-call-collateral-mobilizer, 514-tokenized-fund-collateral-validator. Open at: https://ainumbers.co/tools/505-tokenized-collateral-eligibility-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Classify tokenized assets for DTC/Fed eligibility, Basel HQLA tier, and transfer restrictions. Outputs: dtc_status, hqla_tier, final_haircut_pct, adjusted_value, compliance_flags, mandate_type. Output schema: call describe_tool("check_tokenized_collateral_eligibility").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
checklist_step_receiptCompletes ONE step of a checklist/SOP run and returns its OCG v0.4 step receipt: execution_hash chains to prev_step_receipt_digest (pass the previous step's execution_hash, or omit…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Description, as the server wrote it
Completes ONE step of a checklist/SOP run and returns its OCG v0.4 step receipt: execution_hash chains to prev_step_receipt_digest (pass the previous step's execution_hash, or omit for step 0). A blocking-gate step with evidence_requirement != "none" and no evidence supplied is refused (the caller enforces step order; this tool enforces the evidence requirement per step). Call once per step in order, then pass the full ordered list of returned receipts to checklist_verify_run to check the chain and mint the run receipt.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: false
- openWorldHint: false
checklist_verify_runRecomputes and checks a checklist run: every step receipt's execution_hash, the hash-chain link between consecutive steps, and (if a run_receipt is supplied) the §20.1 RFC 6962 Mer…1 finding
Why it is in this group
- The name "checklist_verify_run" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "checklist_verify_run" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "checklist_verify_run" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Recomputes and checks a checklist run: every step receipt's execution_hash, the hash-chain link between consecutive steps, and (if a run_receipt is supplied) the §20.1 RFC 6962 Merkle root over every step. Returns valid:true/false, per-step ok/hash_ok/link_ok, and broken_at (the zero-based index of the first broken step, or null). Same recompute a human gets from the browser Run Verifier.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
choose_cc_licenseCreative Commons License Chooser: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Creative Commons License Chooser: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-195-creative-commons-license-chooser.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Deterministic two-question decision tree mapping creator answers (waive all rights, allow commercial, allow adaptations) to the matching Creative Commons 4.0 license (CC0, CC BY, BY-SA, BY-ND, BY-NC, BY-NC-SA, BY-NC-ND). Outputs: attribution_required, disclaimer, license_id, license_name, license_url, required_elements, source, spdx_id. Output schema: call describe_tool("choose_cc_license").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_agentic_ai_riskAgentic AI Risk & GPAI Governance Classifier: OpenChainGraph compute node (model_governance).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Agentic AI Risk & GPAI Governance Classifier: OpenChainGraph compute node (model_governance). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-64-ai-act-highrisk-fit-diagnostic. Output feeds: art-04-agent-identity-attestation-checker, art-33-mcp-server-self-attestation-pack, art-62-ap2-payment-receipt-verifier. Open at: https://ainumbers.co/chaingraph/art-67-agentic-ai-risk-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Co-flagship and strongest in-force anchor: classifies autonomy tier and GPAI/systemic-risk obligations (Arts 53-55, IN FORCE since 2 Aug 2025, explicitly unchanged by the Digital Omnibus) for agentic and foundation-model AI in financial services. Outputs: applicable_obligations, autonomy_oversight_verdict, dim_scores, governance_tier, gpai_class, highrisk_interaction, in_force_status, note, overall_grade, overall_score, recommendation. Output schema: call describe_tool("classify_agentic_ai_risk").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_ai_system_governanceAI System Governance Classifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
AI System Governance Classifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-172-ai-risk-impact-assessment-validator. Open at: https://ainumbers.co/chaingraph/art-173-ai-system-governance-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Classify an AI system to its governance tier across EU AI Act (prohibited/high-risk/limited-risk/minimal-risk), NIST AI RMF profile (T1 basic/T2 standard/T3 enhanced), and ISO/IEC 42001 control set (light/standard/enhanced). Outputs: deployment_context, eu_ai_act_tier, gpai_obligations, iso42001_control_set, nist_rmf_profile, use_case. Output schema: call describe_tool("classify_ai_system_governance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_annex3_decisioning_obligationsEU AI Act Annex III FS Decisioning Obligations Classifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
EU AI Act Annex III FS Decisioning Obligations Classifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-64-ai-act-highrisk-fit-diagnostic. Open at: https://ainumbers.co/chaingraph/art-238-classify-annex3-decisioning-obligations.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: EXTENDS run_ai_act_highrisk_fit (art-64). Outputs: all_obligations_met, annex3_category, art12_logging_required, art26_deployer_duties_apply, compliance_gaps, db_registration_required, do_now, enforcement_dates, enforcement_readiness, extends_note, extends_tool, fria_required, is_5b_creditworthiness, is_5c_life_health_insurance, is_high_risk, obligations, regulatory_basis, scope_verdict, table_version. Output schema: call describe_tool("classify_annex3_decisioning_obligations").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_blockchain_quantum_riskBlockchain / Stablecoin Quantum-Risk Classifier: OpenChainGraph compute node (model_governance).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Blockchain / Stablecoin Quantum-Risk Classifier: OpenChainGraph compute node (model_governance). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-85-pqc-timeline-fit-diagnostic, 499-crypto-asset-inventory-classifier. Output feeds: cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-89-blockchain-quantum-risk-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Classifies quantum-exposure risk for blockchain/stablecoin assets: exposed public-key percentage, address reuse, and migration-path maturity (BIP-360/XRPL/Ethereum roadmaps). Outputs: address_reuse_pct, asset_type, bip360_status, exposed_pct, exposure_thresholds, migration_readiness, note, quantum_risk_tier, reference_version, reuse_risk, roadmap_ref, signature_scheme. Output schema: call describe_tool("classify_blockchain_quantum_risk").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_bold_challenge_finalityBoLD Challenge-Window Finality Classifier: OpenChainGraph compute node (settlement_finality_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
BoLD Challenge-Window Finality Classifier: OpenChainGraph compute node (settlement_finality_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-59-settlement-asset-finality-classifier. Open at: https://ainumbers.co/chaingraph/art-321-rhc-bold-finality-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Classifies a settlement-finality claim on Robinhood Chain, an Arbitrum Orbit dedicated blockchain using BoLD interactive fraud proofs, into soft, posted, challengeable, or final. Outputs: challenge_window_seconds, claim_verdict, earliest_final_at, finality_claim, finality_class, verdict. Output schema: call describe_tool("classify_bold_challenge_finality").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_digital_asset_regulatoryDigital Asset Regulatory Classifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Digital Asset Regulatory Classifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: 512-tokenized-security-lifecycle-validator. Open at: https://ainumbers.co/tools/510-digital-asset-regulatory-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Classify a tokenized asset under GENIUS Act, MiCA, MiFID II, and EU DLT Pilot Regime. Outputs: classification_results, compliance_flags, iso20022_party_identification. Output schema: call describe_tool("classify_digital_asset_regulatory").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_dora_incidentDORA Major-Incident Reporting Threshold Classifier: OpenChainGraph compute node (infrastructure_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
DORA Major-Incident Reporting Threshold Classifier: OpenChainGraph compute node (infrastructure_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-29-dora-readiness-diagnostic. Output feeds: pnr-01-dora-ict-cascade-simulator, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-09-dora-incident-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: DORA Article 19/20 reporting determination and reporting-clock start. Outputs: major_incident, determination_code, qualifying_criteria, reporting_clock, entity_type, incident_type, cross_border, third_party_ict, competent_authority_note, regulatory_framework. Output schema: call describe_tool("classify_dora_incident").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_eccn_dual_useECCN / Dual-Use Classifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ECCN / Dual-Use Classifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-90-sanctions-screening-fit-diagnostic. Output feeds: art-95-circumvention-diligence-assessor, cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-94-eccn-dual-use-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Decision-tree from product attributes to ECCN (EAR) + EU Annex I category + controlling regime (Wassenaar/MTCR/AG/NSG) + licence-requirement logic, including 2025 emerging-tech controls (quantum/semiconductor/AM/peptide). Outputs: classification_basis, controlling_regime, eccn, emerging_tech_note, eu_annex_i_category, key_dates, licence_required, note, red_flags, reference_version. Output schema: call describe_tool("classify_eccn_dual_use").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_eudr_commodity_scopeEUDR Commodity Scope Classifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
EUDR Commodity Scope Classifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-166-eudr-geolocation-plot-validator. Open at: https://ainumbers.co/chaingraph/art-167-eudr-commodity-scope-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Classify an HS code against EUDR Annex I to determine commodity scope (cattle, cocoa, coffee, oil palm, rubber, soya, wood, or out-of-scope), operator/trader role, SME status, DDS filing obligation, and enforcement deadline (large/medium 2026-12-30; micro/SME 2027-06-30). Outputs: commodity, dds_filing_required, deadline, due_diligence_required, entity_role, geo_exemption_eligible, hs4_matched, in_scope, is_micro, is_sme, obligations. Output schema: call describe_tool("classify_eudr_commodity_scope").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_ifrs17_measurement_modelIFRS 17 Measurement Model Classifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
IFRS 17 Measurement Model Classifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-178-ifrs17-csm-rollforward-validator. Open at: https://ainumbers.co/chaingraph/art-177-ifrs17-measurement-model-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Classify insurance contracts to their IFRS 17 measurement model: Premium Allocation Approach (PAA) for coverage periods of 12 months or less, Variable Fee Approach (VFA) for direct-participating contracts with investment-linked features (not reinsurance), or General Measurement Model/BBA as the d... Outputs: coverage_period_months, direct_participating, eligible_models, is_reinsurance, measurement_model, paa_eligible, vfa_eligible. Output schema: call describe_tool("classify_ifrs17_measurement_model").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_nis2_entityNIS2 Entity Scope Classifier (Essential / Important / Out-of-Scope): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
NIS2 Entity Scope Classifier (Essential / Important / Out-of-Scope): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-142-nis2-art21-gap-checker. Open at: https://ainumbers.co/chaingraph/art-141-nis2-entity-scope-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Classify an entity as Essential, Important, or Out-of-Scope under NIS2 Directive 2022/2555 Annex I and II, applying sector codes, employee/turnover size thresholds, and automatic-essential carve-outs (DNS providers, qualified trust service providers, public electronic communications networks). Outputs: annex, annual_turnover_eur, applicable_penalties, automatic_essential, classification_basis, employee_count, entity_classification, sector_code. Output schema: call describe_tool("classify_nis2_entity").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_qm_apr_apor_spreadQM APR-APOR Spread Classifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
QM APR-APOR Spread Classifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-218-qm-points-and-fees. Open at: https://ainumbers.co/chaingraph/art-219-qm-apr-apor-spread.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: QM APR-APOR spread test per Reg Z §1026.43(e)(2)(vi) and §1026.43(b)(4). Outputs: apor_pct, applicable_threshold_pct, apr_pct, fr_citation, general_qm_pass, headroom_pct, hpct_threshold_pct, is_hpct, is_manufactured_housing, lien_type, loan_amount, note, qm_status, regulatory_basis, spread_pct, threshold_basis, year. Output schema: call describe_tool("classify_qm_apr_apor_spread").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_sco60_exposureSCO60 Crypto-Asset Exposure Classifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
SCO60 Crypto-Asset Exposure Classifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-281-sco60-crypto-asset-exposure-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Classifies a crypto-asset position into Basel SCO60 Group 1a, 1b, 2a, or 2b (BCBS d545 Prudential treatment of cryptoasset exposures), applies the Group 1 infrastructure-risk capital add-on, and checks the Group 2 exposure limit of 1% of Tier 1 capital. Outputs: base_risk_weight_pct, classification, gaps, group, group2_exposure_pct_tier1, group2_limit_breached, infra_addon_capped, infra_addon_pct_applied, infra_addon_pct_input, pillar3_precheck_pass, risk_weight_applied_pct. Output schema: call describe_tool("classify_sco60_exposure").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_settlement_asset_finalitySettlement-Asset & Legal-Finality Classifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Settlement-Asset & Legal-Finality Classifier: OpenChainGraph compute node (compliance_mandate). Regulatory deadline: 2026-Q3 (ECB Pontes pilot end-Q3 2026; DTCC Collateral AppChain production Oct 2026. Verify SFD/PFMI designations against current regulatory status before that date.). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-56-tokenized-settlement-fit-diagnostic. Output feeds: art-58-cross-network-settlement-validator, 506-onchain-cash-leg-finality-checker, 510-digital-asset-regulatory-classifier. Open at: https://ainumbers.co/chaingraph/art-59-settlement-asset-finality-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Classifies the settlement asset (CBM token / tokenized commercial bank deposit / regulated stablecoin / e-money token) against its legal-finality regime (EU SFD 98/26/EC / CPMI-IOSCO PFMI Principle 8 / UCC Article 12 control) -> finality tier 1-4 + singleness-of-money verdict. Outputs: applicable_regime, finality_gaps, finality_tier, note, recommendation, settlement_asset_class, singleness_verdict, status_asof. Output schema: call describe_tool("classify_settlement_asset_finality").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
classify_vida_platform_deemed_supplierViDA Platform Deemed Supplier Classifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ViDA Platform Deemed Supplier Classifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-163-vida-oss-registration-router. Open at: https://ainumbers.co/chaingraph/art-162-vida-platform-deemed-supplier-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compile_work_mandateWork Mandate Compiler: OpenChainGraph compute node (governance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Work Mandate Compiler: OpenChainGraph compute node (governance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-274-compile-work-mandate.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Compiles a §22 Work Mandate document into a deterministic §21.4 gated-chain config. Outputs: chain_config. Output schema: call describe_tool("compile_work_mandate").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compose_eval_attestation_receiptEval Attestation Receipt Composer: OpenChainGraph compute node (governance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Eval Attestation Receipt Composer: OpenChainGraph compute node (governance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-438-eval-attestation-receipt-composer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Hashes a third-party eval log (e.g. Outputs: attestation_determination, claim_strength, eval_format, eval_id, eval_log_hash, mandate_reference, not_proven, verify_instructions. Output schema: call describe_tool("compose_eval_attestation_receipt").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_15c3_3_reserve15c3-3 Customer Reserve Formula Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
15c3-3 Customer Reserve Formula Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-397-lint-trace-cat-reports. Open at: https://ainumbers.co/chaingraph/art-396-compute-15c3-3-reserve.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: SEC Rule 15c3-3 Exhibit A customer reserve formula: credit items (customer free credit balances, margin credit balances, payables) against allowable debit items (margin-account debits with the 1% collateral haircut, failed-to-deliver debits with the 30-day aging exclusion, other allowable debits)... Outputs: credit_items, debit_items, deposit_sufficient, note, pab_variant, regulatory_basis, reserve_account_balance_musd, reserve_requirement_musd, rules_version, surplus_shortfall_musd, total_credits_musd, total_debits_musd. Output schema: call describe_tool("compute_15c3_3_reserve").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_aca_affordability_safe_harborACA Affordability Safe-Harbor Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ACA Affordability Safe-Harbor Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-299-aca-esrp-exposure. Open at: https://ainumbers.co/chaingraph/art-298-aca-affordability-safe-harbor.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Compute the ACA employer-mandate affordability percentage under each of the three IRC 4980H(a)(1)(B) safe harbors (W-2 wages, rate-of-pay, federal poverty line) against a supplied lowest-cost self-only monthly premium, using the version-pinned 2026 affordability threshold (9.96%, IRS Rev. Outputs: affordability_pct, error, harbors, harbors_satisfied, lowest_cost_self_only_monthly_premium, satisfies_any_harbor, tax_year. Output schema: call describe_tool("compute_aca_affordability_safe_harbor").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_annuityAnnuity PV / FV / Payment Solver: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Annuity PV / FV / Payment Solver: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-327-tvm-annuity.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Solves present value, future value, or payment for an ordinary annuity or annuity-due, given the other two plus rate and number of periods, using the standard closed-form annuity factor. Outputs: annuity_factor, due, fv, note, nper, pmt, pv, rate_pct, regulatory_basis, solved_for. Output schema: call describe_tool("compute_annuity").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_basel_haircut_adjusted_exposureCollateral Haircut Engine (Basel CRE22): OpenChainGraph compute node (regulatory_reporting).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Collateral Haircut Engine (Basel CRE22): OpenChainGraph compute node (regulatory_reporting). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-444-collateral-haircut-engine.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Basel CRE22 comprehensive-approach collateral haircut engine for counterparty credit risk: applies a caller-supplied, versioned supervisory haircut table (policy input, not hardcoded) to each collateral item's asset class/maturity bucket, scales for a non-standard holding period via the CRE22.68... Outputs: collateral_adjusted_total, collateral_items, exposure_adjusted, exposure_amount, exposure_asset_class, exposure_currency, exposure_haircut_scaled_pct, haircut_table_version, holding_period_days, item_count, net_exposure, note, override_count, override_missing_reason_count, time_scale_factor, unclassified_count. Output schema: call describe_tool("compute_basel_haircut_adjusted_exposure").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_basel31_deltaBasel 3.1 Reporting Delta Calculator: OpenChainGraph compute node (capital_assessment).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Basel 3.1 Reporting Delta Calculator: OpenChainGraph compute node (capital_assessment). Regulatory deadline: 2027-01-01 (UK PRA PS1/26 Basel 3.1 go-live — January 1, 2027). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: sim-03-basel-rwa-scenario-modeler, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-07-basel31-reporting-delta-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_bond_durationBond Macaulay / Modified Duration: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Bond Macaulay / Modified Duration: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-330-tvm-dv01, art-331-tvm-convexity. Open at: https://ainumbers.co/chaingraph/art-329-tvm-bond-duration.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Macaulay and modified duration for a standard even-period bullet bond, given face value, coupon rate, yield to maturity, years to maturity, and compounding frequency. Outputs: coupon_rate_pct, day_count_convention, face_value, macaulay_duration_years, modified_duration_years, note, num_periods, periods_per_year, price, regulatory_basis, years_to_maturity, ytm_pct. Output schema: call describe_tool("compute_bond_duration").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_breakevenBreakeven / CVP Analysis: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Breakeven / CVP Analysis: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-328-tvm-breakeven.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Standard cost-volume-profit breakeven analysis: breakeven units and revenue from fixed costs, price per unit, and variable cost per unit, plus contribution margin ratio and an optional margin-of-safety calculation against a supplied current volume. Outputs: breakeven_revenue, breakeven_units, contribution_margin_ratio, fixed_costs, margin_of_safety_pct, margin_of_safety_units, note, price_per_unit, regulatory_basis, unit_contribution, variable_cost_per_unit. Output schema: call describe_tool("compute_breakeven").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_canton_app_reward_estimateCanton App-Reward Estimator (CIP-0104): OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Canton App-Reward Estimator (CIP-0104): OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-392-compute-canton-app-reward-estimate.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Estimates a Canton Network app provider's Canton Coin reward for one round under CIP-0104 (approved 2026-02-12): app rewards are proportional to confirmed envelope bytes (traffic where the app-provider party appears as confirmer) against that round's app-reward pool share of the minting curve. Outputs: app_reward_pool_cc, app_reward_pool_share, cc_reward_estimate, confirmed_envelope_bytes, confirmed_share_of_traffic, disambiguation, pool_share_source, protocol_version, round_total_envelope_bytes, round_total_mint_cc. Output schema: call describe_tool("compute_canton_app_reward_estimate").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_canton_traffic_costCanton Synchronizer Traffic-Cost Calculator: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Canton Synchronizer Traffic-Cost Calculator: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-391-compute-canton-traffic-cost.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes Canton Network synchronizer traffic cost (CIP-0042/CIP-0084 regime): fee = message megabytes x the Tokenomics-Committee-set USD/MB rate, converted to Canton Coin burned at a caller-supplied CC/USD price. Outputs: cc_burned, cc_usd_price, disambiguation, effective_rate_usd_per_mb, envelope_mb, free_period_applies, free_period_days, free_period_source, is_transfer_preapproval, preapproval_age_days, protocol_version, rate_source, rate_usd_per_mb, usd_traffic_cost. Output schema: call describe_tool("compute_canton_traffic_cost").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_cdd_ownership_25pctFinCEN CDD 25% Beneficial Ownership Attribution: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
FinCEN CDD 25% Beneficial Ownership Attribution: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-269-validate-w8-series-structural. Open at: https://ainumbers.co/chaingraph/art-268-compute-cdd-ownership-25pct.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Recursive indirect natural-person beneficial ownership computation via ownership-tier multiplication. Outputs: below_threshold, below_threshold_count, beneficial_owner_count, beneficial_owners, entities_evaluated, is_beneficial_owner, methodology, not_legal_advice, pii_note, regime_note, regulatory_basis, table_source, table_version, threshold_pct. Output schema: call describe_tool("compute_cdd_ownership_25pct").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_convexityBond Convexity: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Bond Convexity: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-329-tvm-bond-duration. Open at: https://ainumbers.co/chaingraph/art-331-tvm-convexity.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Standard closed-form convexity for a bullet bond, annualized by compounding frequency squared. Outputs: convexity, convexity_price_adjustment_pct, coupon_rate_pct, face_value, note, num_periods, periods_per_year, price, regulatory_basis, years_to_maturity, ytm_pct. Output schema: call describe_tool("compute_convexity").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_counterparty_limit_checkCounterparty Internal Limit Check: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Counterparty Internal Limit Check: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-446-counterparty-internal-limit-check.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Counterparty internal credit-limit check: compares each counterparty's caller-supplied current exposure against its board-approved internal limit line (settlement, pre-settlement/PFE, or aggregate limit type), computes utilization percent and headroom, and flags each counterparty WITHIN_LIMIT, WA... Outputs: breach_list, counterparties, note, regulatory_basis, warning_list. Output schema: call describe_tool("compute_counterparty_limit_check").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_cross_border_feesCross-Border B2B Fee Calculator: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Cross-Border B2B Fee Calculator: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-367-compute-cross-border-fees.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Itemizes a single cross-border B2B invoice's total cost stack: FX spread cost, payment-method/correspondent fee, VAT or reverse-charge cost, documentary-credit cost, and reconciliation cost, plus total cost as a percentage of invoice value. Outputs: dest_country, disambiguation, doc_cost, fx_cost, fx_spread_bps, invoice_amount, method_fee, origin_country, pct_of_invoice, recon_cost, regulatory_basis, total_cost, vat_cost, vat_rate. Output schema: call describe_tool("compute_cross_border_fees").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_cyber_incident_notification_clockCyber Incident Notification Clock: OpenChainGraph compute node (attestation_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Cyber Incident Notification Clock: OpenChainGraph compute node (attestation_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-428-cyber-incident-clock.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Starts three parallel regulatory notification-deadline clocks from one hash-anchored cyber-incident determination timestamp: the 36-hour interagency banking-regulator rule (12 CFR 53/225/304), the 4-business-day SEC Form 8-K Item 1.05 notification (weekends-only business-day arithmetic; no federa... Outputs: determination_at, determination_at_parsed, determination_evidence_hash, determination_evidence_hash_well_formed, determinations, evaluated_at, incident_id, note. Output schema: call describe_tool("compute_cyber_incident_notification_clock").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_discount_window_capacityDiscount Window Borrowing-Capacity Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Discount Window Borrowing-Capacity Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-427-discount-window-capacity.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Federal Reserve Discount Window borrowing-capacity calculator: lendable value = sum of pledged collateral positions x published Fed collateral margins (margin table effective date is caller-supplied policy input, not hardcoded) compared against a runnable-liability / uninsured-deposit coverage ta... Outputs: capacity_compliant, capacity_surplus_shortfall_musd, collateral_par_value_musd, coverage_pct, coverage_target_pct, lendable_value_musd, margin_table_version, note, runnable_liabilities_musd. Output schema: call describe_tool("compute_discount_window_capacity").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_disparity_metricsCompute Disparate Impact Metrics: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Compute Disparate Impact Metrics: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-230-compute-hmda-rate-spread. Open at: https://ainumbers.co/chaingraph/art-229-compute-disparity-metrics.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes three fair lending disparate impact metrics from aggregate lending counts: 4/5ths (80%) rule (adverse_impact_ratio) per EEOC 29 CFR §1607.4(D), z-statistic for statistical significance of the approval-rate difference, and standardised mean difference (Cohen's d). Outputs: adverse_impact_ratio, four_fifths_flag, four_fifths_result, four_fifths_threshold, group_a, group_b, n_total, odds_ratio, pii_note, pooled_proportion, regulatory_basis, standardized_mean_difference, table_source, table_version, two_proportion_z, z_critical_flag_onetail_05, z_critical_flag_twotail_05, zero_pii_confirmation. Output schema: call describe_tool("compute_disparity_metrics").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_dscrDSCR & Interest Coverage Ratio Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
DSCR & Interest Coverage Ratio Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-363-compute-dscr.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Debt Service Coverage Ratio and Interest Coverage Ratio suite: Basic/Cash/FCF DSCR, Fixed Charge Coverage Ratio (FCCR), EBIT- and EBITDA-basis ICR, and Net/Gross Leverage, evaluated against a published lender threshold matrix (investment grade, leveraged loan, CRE, infrastructure, SME/mid-market). Outputs: basic_dscr, cash_dscr, fccr, fcf_dscr, gross_leverage, icr_ebit_basis, icr_ebitda_basis, lender_assessment, net_leverage, note, regulatory_basis, total_debt_service_musd. Output schema: call describe_tool("compute_dscr").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_dti_ratiosDTI Ratio Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
DTI Ratio Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-222-agency-eligibility-matrix. Open at: https://ainumbers.co/chaingraph/art-335-compute-dti-ratios.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Front-end (housing) and back-end (total) debt-to-income ratios per Fannie Mae Selling Guide B3-6-02 and Freddie Mac Single-Family Seller/Servicer Guide 5401.2. Outputs: back_end_dti_pct, dti_tier, front_end_dti_pct, gross_monthly_income, housing_payment_pitia, max_dti_pct, note, other_monthly_debts, regulatory_basis, total_monthly_debt, underwriting_type, within_max. Output schema: call describe_tool("compute_dti_ratios").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_dv01Bond DV01 (Price Value of a Basis Point): OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Bond DV01 (Price Value of a Basis Point): OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-329-tvm-bond-duration. Open at: https://ainumbers.co/chaingraph/art-330-tvm-dv01.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: DV01 / price value of a basis point for a standard even-period bullet bond, computed by full central-difference reprice at yield plus and minus a declared basis-point shock, not the linear modified-duration approximation. Outputs: coupon_rate_pct, dv01, face_value, method, note, periods_per_year, price, price_down_shock, price_up_shock, regulatory_basis, shock_size_bp, years_to_maturity, ytm_pct. Output schema: call describe_tool("compute_dv01").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_escrow_analysisRESPA Aggregate Escrow Analysis: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
RESPA Aggregate Escrow Analysis: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-235-test-hpml-escrow. Open at: https://ainumbers.co/chaingraph/art-342-compute-escrow-analysis.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 12 CFR 1024.17 (Reg X) aggregate escrow accounting method: builds a 12-month trial running balance from a starting balance, a monthly escrow deposit, and the projected annual disbursement schedule, compares the low point to the 1/6-of-annual-disbursements cushion target, and classifies the accoun... Outputs: account_status, cushion_fraction_used, cushion_target, deficiency_amount, low_point_balance, low_point_month, monthly_deficiency_spread_amount, monthly_escrow_payment, monthly_shortage_spread_amount, new_monthly_escrow_payment, note, regulatory_basis, shortage_amount, shortage_spread_required, spread_vs_target, starting_balance, surplus_amount, surplus_refund_required, total_annual_disbursements, trial_balances. Output schema: call describe_tool("compute_escrow_analysis").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_esrp_exposureACA Employer Shared Responsibility Payment Exposure Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ACA Employer Shared Responsibility Payment Exposure Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-298-aca-affordability-safe-harbor. Output feeds: art-300-aca-226j-response-evidence-pack. Open at: https://ainumbers.co/chaingraph/art-299-aca-esrp-exposure.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Compute proposed IRC 4980H(a) ("no offer to 95%") and 4980H(b) ("unaffordable / not minimum value") Employer Shared Responsibility Payment exposure from supplied full-time-employee counts, minimum-essential-coverage offer counts, and PTC-triggering employee counts, using the version-pinned 2026 p... Outputs: a_applicable, a_exposure_annual, a_monthly_per_employee, b_applicable, b_exposure_annual, b_monthly_per_employee, controlling_exposure_annual, controlling_penalty, coverage_offer_rate, error, offer_rate_threshold, tax_year. Output schema: call describe_tool("compute_esrp_exposure").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_experience_modNCCI Experience Modification Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
NCCI Experience Modification Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-346-compute-experience-mod.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Workers'-compensation experience rating modification (NCCI Experience Rating Plan Manual published national formula): per-claim primary/excess loss split at the state split point, then Mod = (Ap + W×Ae + (1-W)×Ee + B) / (Ep + B). Outputs: actual_excess_losses, actual_primary_losses, actual_total_losses, ballast_value, claim_count, expected_excess_losses, expected_losses, expected_primary_losses, mod, note, rating_class, regulatory_basis, split_point, weighting_value. Output schema: call describe_tool("compute_experience_mod").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_fdic_assessment_rateFDIC Deposit-Insurance Assessment Rate Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
FDIC Deposit-Insurance Assessment Rate Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-431-fdic-assessment-rate-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: FDIC deposit-insurance assessment rate calculator (12 CFR 327): looks up the base assessment rate for a supplied composite CAMELS + financial-ratio score against a caller-supplied rate schedule, then applies the unsecured-debt and brokered-deposit adjustments and floors/caps the result to a calle... Outputs: assessment_base_musd, base_rate_bp, brokered_deposit_adjustment_bp, estimated_quarterly_assessment_musd, note, npr_banner, rate_cap_bp, rate_floor_bp, rate_floor_or_cap_applied, rate_schedule_version, total_rate_bp, total_score, unsecured_debt_adjustment_bp. Output schema: call describe_tool("compute_fdic_assessment_rate").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_federal_withholdingFederal Withholding Calculator (Percentage Method): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Federal Withholding Calculator (Percentage Method): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-339-compute-gross-to-net. Open at: https://ainumbers.co/chaingraph/art-338-compute-federal-withholding.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Federal income tax withholding via the IRS Publication 15-T percentage method, Worksheet 1A, for a 2020-or-later Form W-4. Outputs: adjusted_annual_wage_amount, bracket_at_least, bracket_rate, constants_version, error, federal_withholding_per_period, filing_status, note, pay_frequency, periods_per_year, regulatory_basis, step3_credit_this_period, step4c_extra_withholding_per_period, supported_tax_years, tax_year, tentative_annual_withholding, tentative_withholding_this_period. Output schema: call describe_tool("compute_federal_withholding").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_fha_mip_eligibilityFHA MIP Eligibility Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
FHA MIP Eligibility Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-224-fha-mip-eligibility.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: FHA mortgage insurance premium (MIP) eligibility and cost calculator per HUD Handbook 4000.1. Outputs: annual_mip, dti, fha_eligible, fico_eligible, ltv_eligible, max_ltv_pct, note, pii_note, regulatory_basis, table_source, table_version, ufmip. Output schema: call describe_tool("compute_fha_mip_eligibility").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_flsa_regular_rateFLSA Regular Rate & Overtime Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
FLSA Regular Rate & Overtime Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-340-compute-flsa-regular-rate.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: FLSA regular rate of pay and overtime premium per 29 CFR 778, Subpart C, including nondiscretionary-bonus reallocation into the regular rate (778.110, 778.208-778.209) and the 0.5x overtime premium for hours over 40 in the workweek (778.107). Outputs: constants_version, discretionary_bonus_excluded, hourly_rate, hours_worked_week, nondiscretionary_bonus_amount, note, other_includable_pay, overtime_hours, overtime_premium_pay, regular_rate, regulatory_basis, straight_time_pay, total_pay_due, total_remuneration. Output schema: call describe_tool("compute_flsa_regular_rate").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_fr2052a_inflow_outflow_classificationFR 2052a Inflow/Outflow Bucket Classifier: OpenChainGraph compute node (regulatory_reporting).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
FR 2052a Inflow/Outflow Bucket Classifier: OpenChainGraph compute node (regulatory_reporting). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-437-fr2052a-inflow-outflow-classifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: FR 2052a complex-institution liquidity monitoring filing-layer kernel, scoped to the inflow/outflow section: product/maturity-bucket classification against a caller-supplied Appendix IV-style boundary table (versioned policy input, not hardcoded), intercompany elimination (excluded from external... Outputs: boundary_table_version, elimination_total_musd, form_2052a, note, override_count, override_missing_reason_count, row_count, rows, total_inflow_musd, total_net_musd, total_outflow_musd. Output schema: call describe_tool("compute_fr2052a_inflow_outflow_classification").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_fund_expense_ratiosCompute Fund Expense Ratios: OpenChainGraph compute node (attestation_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Compute Fund Expense Ratios: OpenChainGraph compute node (attestation_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-375-compute-fund-expense-ratios.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes a fund's gross and net expense ratios and Total Expense Ratio (TER) from SUPPLIED gross expense components (flat amounts or accruals, using the identical accrual conventions as recompute_fund_nav/FN-1) and average net assets, applying any declared fee waivers/caps strictly in a DECLARED... Outputs: components, fence, fund_id, gross_expense_ratio, net_expense_ratio, not_proven, period_end, period_start, regulatory_framework, rounding, structural_error, total_expense_ratio. Output schema: call describe_tool("compute_fund_expense_ratios").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_fx_netting_positionsMultilateral FX Netting Calculator: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Multilateral FX Netting Calculator: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-368-compute-fx-netting-positions.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Multilateral FX netting across up to 8 currencies: nets each currency's payable/receivable exposures in FCY, converts to USD at a caller-supplied spot-plus-forward-points effective rate, and returns gross volume, net volume, netting efficiency, estimated settlement savings, and per-currency resid... Outputs: currency_count, estimated_settlement_savings_usd, gross_volume_usd, net_volume_usd, netting_efficiency_pct, positions, regulatory_basis. Output schema: call describe_tool("compute_fx_netting_positions").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_gross_to_netGross-to-Net Payroll Calculator (FICA): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Gross-to-Net Payroll Calculator (FICA): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-338-compute-federal-withholding. Open at: https://ainumbers.co/chaingraph/art-339-compute-gross-to-net.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Gross-to-net payroll calculation: FICA (Social Security 6.2% up to the OASDI contribution and benefit base for the selected tax year, Medicare 1.45% uncapped, Additional Medicare Tax 0.9% on cumulative wages over $200,000 per IRC 3101(b)(2)), pretax deduction ordering (401(k)/HSA/Section 125 redu... Outputs: additional_medicare_tax, additional_medicare_threshold, additional_medicare_wages_this_period, constants_version, error, federal_withholding_per_period, fica_tax_total, fica_wages_this_period, medicare_tax, net_pay, note, post_tax_other_deductions, pretax_reduces_fica_and_fit, regulatory_basis, social_security_tax, ss_taxable_wages_this_period, ss_wage_base, supported_tax_years, tax_year, ytd_fica_wages_after_period. Output schema: call describe_tool("compute_gross_to_net").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_hmda_rate_spreadCompute HMDA Rate Spread: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Compute HMDA Rate Spread: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-229-compute-disparity-metrics. Open at: https://ainumbers.co/chaingraph/art-230-compute-hmda-rate-spread.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes the HMDA rate spread (APR minus APOR) per FFIEC methodology and classifies against HMDA reportability thresholds: 1.5 pp (first lien), 3.5 pp (subordinate lien), 6.5 pp (HELOC). Outputs: apor_pct, apor_source_note, apr_pct, hmda_report_code, is_reportable, lien_type, lock_date, pii_note, product_type, rate_spread_pct, regulatory_basis, reportability_threshold_pct, table_source, table_version. Output schema: call describe_tool("compute_hmda_rate_spread").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_ipfs_cidIPFS CID Computer: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
IPFS CID Computer: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-210-ipfs-cid-computer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes a deterministic IPFS CIDv1 (SHA-256 multihash, multibase base32) for any text input using a pure-JavaScript SHA-256 implementation. Outputs: byte_length, cid, cid_bytes_hex, codec, codec_code, digest_hex, digest_length, disclaimer, known_vector_verified, multihash_fn, multihash_fn_code. Output schema: call describe_tool("compute_ipfs_cid").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_irrInternal Rate of Return (IRR): OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Internal Rate of Return (IRR): OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-325-tvm-irr.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Internal rate of return for an equal-period cash flow series, solved by deterministic bisection over a declared rate bracket with declared tolerance and iteration cap. Outputs: bracket_hi_pct, bracket_lo_pct, converged, irr_pct, iterations, method, note, num_cash_flows, regulatory_basis, tolerance. Output schema: call describe_tool("compute_irr").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_large_exposures_limitLarge Exposures Limit Check: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Large Exposures Limit Check: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-425-large-exposures-limit-check.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Basel III large exposures framework (BCBS 283) and U.S. Outputs: breach_list, caller_is_gsib, counterparties, groups, note, regulatory_basis, tier1_capital_musd. Output schema: call describe_tool("compute_large_exposures_limit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_lcm_rate_derivationLCM Rate Derivation Calculator: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
LCM Rate Derivation Calculator: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-255-compute-lcm-rate-derivation.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes the Loss Cost Multiplier (LCM) and indicated insurance rate from user-supplied loss costs and expense/profit loadings. Outputs: complement_loss_cost, credibility_weighted_loss_cost, credibility_z, current_rate, denominator_valid, fixed_expense_pct, indicated_rate, issues, lae_pct, lcm, lcm_denominator, not_legal_advice, pii_note, profit_pct, proprietary_data_notice, pure_loss_cost, pure_loss_cost_loaded, rate_change_direction, rate_change_pct, regulatory_basis, table_source, table_version, total_loading, variable_exp_pct. Output schema: call describe_tool("compute_lcm_rate_derivation").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_lcr_nsfr_leverageLCR / NSFR / Leverage Ratio Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
LCR / NSFR / Leverage Ratio Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-364-compute-lcr-nsfr-leverage.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Basel III Liquidity Coverage Ratio (BCBS 238), Net Stable Funding Ratio (BCBS 295), and Leverage Ratio (BCBS 270, finalized BCBS 360) point-in-time calculator from caller-supplied HQLA positions, outflow/inflow categories, ASF/RSF items, and capital/exposure figures. Outputs: lcr, leverage, note, nsfr, regulatory_basis. Output schema: call describe_tool("compute_lcr_nsfr_leverage").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_llpa_stackLLPA Stack Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
LLPA Stack Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-222-agency-eligibility-matrix. Open at: https://ainumbers.co/chaingraph/art-221-llpa-stack.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Fannie Mae public LLPA (Loan-Level Price Adjustment) matrix calculator. Outputs: base_llpa, components, feature_llpa, fico_band, fthb_eligible, fthb_waiver, ltv_band, pii_note, regulatory_basis, table_source, table_version, total_llpa_pct. Output schema: call describe_tool("compute_llpa_stack").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_ltv_ratiosLTV/CLTV/HCLTV Ratio Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
LTV/CLTV/HCLTV Ratio Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-222-agency-eligibility-matrix. Open at: https://ainumbers.co/chaingraph/art-336-compute-ltv-ratios.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Loan-to-value, combined LTV, and home-equity combined LTV per Fannie Mae Selling Guide B2-1.1-03 and Freddie Mac Single-Family Seller/Servicer Guide 5401.1. Outputs: appraised_value, cltv_pct, first_lien_amount, hcltv_pct, heloc_credit_limit, ltv_pct, note, regulatory_basis, sales_price, subordinate_lien_amount, transaction_type, value_used. Output schema: call describe_tool("compute_ltv_ratios").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_mla_maprCompute MLA MAPR (closed-end): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Compute MLA MAPR (closed-end): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-232-compute-scra-rate-cap. Open at: https://ainumbers.co/chaingraph/art-231-compute-mla-mapr.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes the Military Annual Percentage Rate for CLOSED-END consumer credit and tests it against the 36 percent limit that 10 USC 987(b) imposes and 32 CFR 232.4(b) sets out. Outputs: amount_advanced, amount_financed_mapr, application_fee_carve_out_applied, application_fee_carve_out_predicates, approaching_cap_threshold_basis, approaching_cap_threshold_pct, bona_fide_exclusion_available, bona_fide_fee_claimed_total, bracketed, charge_breakdown, converged, credit_class, exceeds_cap, finance_charge_in_schedule, in_scope, iterations, mapr_cap_pct, mapr_determined, mapr_pct, max_payment_count, method_basis, payment_amount, payment_count, payment_count_exceeds_limit, payment_structure, payment_total, periodic_rate, pii_note, prepaid_includable_charges, regulatory_basis, scope_note, table_source, table_version, term_days, total_excluded_charges, total_includable_charges, unit_periods_per_year. Output schema: call describe_tool("compute_mla_mapr").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_mlr_rebateMLR Rebate Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
MLR Rebate Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-344-compute-mlr-rebate.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Medical Loss Ratio numerator/denominator, credibility-adjustment tier, 3-year premium-weighted averaging, and rebate math per 45 CFR 158 (ACA MLR rule). Outputs: adjusted_earned_premium, adjusted_incurred_claims, constants_version, credibility_adjustment_pct_points, credibility_tier, current_year_adjusted_mlr_pct, de_minimis, market, member_life_years, note, raw_mlr_pct, rebate_amount, rebate_owed, rebate_pct_points, regulatory_basis, reporting_year, three_yr_average_mlr_pct, threshold_pct, years_included_in_average. Output schema: call describe_tool("compute_mlr_rebate").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_multilateral_nettingMultilateral Cash Netting: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Multilateral Cash Netting: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-260-allocate-ihb-interest. Open at: https://ainumbers.co/chaingraph/art-259-compute-multilateral-netting.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes N-entity corporate cash netting: gross inter-company positions to net positions to minimum settlement legs using the BIS CPMI greedy matching algorithm (Net Settlement Framework 2012). Outputs: anchor_surface, base_currency, entity_count, entity_net_positions, gross_count, gross_volume, net_count, net_volume, netting_efficiency_pct, not_legal_advice, pii_note, regulatory_basis, settlement_legs, table_source, table_version, wire_count_savings, wire_count_savings_pct. Output schema: call describe_tool("compute_multilateral_netting").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_npvNet Present Value (NPV): OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Net Present Value (NPV): OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-324-tvm-npv.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Net present value of a cash flow series, discounted at a declared periodic rate. Outputs: day_count_convention, discount_rate_pct, mode, note, npv, num_cash_flows, regulatory_basis, total_undiscounted. Output schema: call describe_tool("compute_npv").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_oprisk_sma_2026Basel Operational Risk SMA (2026 Reproposal): OpenChainGraph compute node (capital_assessment).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Basel Operational Risk SMA (2026 Reproposal): OpenChainGraph compute node (capital_assessment). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-356-compute-oprisk-sma-2026.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Basel Standardized Measurement Approach (SMA) for operational-risk capital per the July 2026 US Basel Endgame reproposal (comment period closed 2026-06-18, final expected ~Q4 2026): marginal Business Indicator Component across the $1bn/$30bn buckets (12%/15%/18%) and the Internal Loss Multiplier,... Outputs: average_annual_loss, bucket, business_indicator, business_indicator_component, constants_version, fc_avg, ildc_avg, internal_loss_multiplier, loss_component, note, operational_risk_capital, regulatory_basis, rule_status, rwa, sc_avg, use_us_ilm_neutralization. Output schema: call describe_tool("compute_oprisk_sma_2026").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_perp_fundingPerp Funding and Carry Calculator: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Perp Funding and Carry Calculator: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-270-perp-funding-carry.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes perpetual futures funding rates, compound annual APR, and cross-venue funding differential arbitrage (Hyperliquid hourly vs Binance 8-hour cadence). Outputs: annualized_basis_pct, asset, basis_pct, basis_perp_price, basis_spot_price, basis_usd, cadence_hours, compound_annualized_rate_pct, cross_venue_arb, delta_neutral_carry, funding_rate_pct_per_period, hl_funding_cap_pct_per_hour, hl_typical_rate_note, hourly_rate_pct, not_financial_advice, pii_note, regulatory_basis, simple_annualized_rate_pct, table_source, table_version, venue. Output schema: call describe_tool("compute_perp_funding").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_perp_marginPerp Margin and Liquidation Calculator: OpenChainGraph compute node (derivatives_margin_health).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Perp Margin and Liquidation Calculator: OpenChainGraph compute node (derivatives_margin_health). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-214-perp-position-lifecycle. Open at: https://ainumbers.co/chaingraph/art-213-perp-liquidation-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes liquidation price and margin health for perpetual futures positions. Outputs: buffer, buffer_pct, cross_margin_efficiency, disclaimer, distance_to_liq_pct, entry_price, health, imr_pct, initial_margin, leverage, liq_price, maintenance_threshold, margin_balance, mark_price, mark_price_note, mmr_pct, mode, notional, portfolio_margin_note, position_size, side, unrealized_pnl, venue. Output schema: call describe_tool("compute_perp_margin").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_pqc_deadline_ladderCNSA 2.0 Deadline Ladder Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
CNSA 2.0 Deadline Ladder Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-387-pqc-deadline-ladder-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Per-row CNSA 2.0 post-quantum migration deadline for a supplied system inventory (system class, asset type, deployment date): applicable deadline, days remaining, the earliest binding constraint, and a FIPS 140-2 Historical-list exposure flag. Outputs: note, policy_deadlines_used, reference_date, rows, summary. Output schema: call describe_tool("compute_pqc_deadline_ladder").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_pt_yt_yieldPendle Yield Tokenization Analyzer (PT/YT): OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Pendle Yield Tokenization Analyzer (PT/YT): OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-273-pendle-yield.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Decomposes Pendle Finance yield tokenization: PT implied fixed yield, YT leverage and break-even APY, PT+YT=1 invariant check, and time-to-maturity analytics. Outputs: days_to_maturity, invariant_holds, investment_usd, not_financial_advice, pii_note, pt_discount_pct, pt_implied_fixed_yield_pct, pt_price, pt_profit_at_maturity_usd, pt_return_to_maturity_pct, pt_simple_apr_pct, pt_units_bought, pt_yt_sum, regulatory_basis, table_source, table_version, underlying_apy_pct, years_to_maturity, yt_break_even_apy_pct, yt_leverage, yt_levered_apy_pct, yt_margin_of_safety_pct, yt_net_pnl_if_underlying_stable_usd, yt_payout_if_underlying_stable_usd, yt_price, yt_profitable_at_current, yt_units_bought. Output schema: call describe_tool("compute_pt_yt_yield").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_raroc_loan_priceRAROC Loan Pricing Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
RAROC Loan Pricing Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-362-compute-raroc-loan-price.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Risk-Adjusted Return on Capital (RAROC) loan pricing per Basel II BCBS 128 (2006) / Basel III BCBS 189 (2010) simplified public approximation of the IRB economic-capital formula (single-factor Vasicek model at 99.9% confidence, or the SA risk-weight bucket table). Outputs: break_even_gap_bps, break_even_spread_bps, capital_approach, drawn_musd, economic_capital_musd, expected_loss_musd, funding_cost_musd, gross_revenue_musd, hurdle_rate_pct, net_income_after_tax_musd, net_income_before_tax_musd, note, operating_cost_musd, raroc_pct, regulatory_basis, undrawn_musd, value_creating, value_spread_pct. Output schema: call describe_tool("compute_raroc_loan_price").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_rbc_action_levelNAIC RBC Action Level Calculator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
NAIC RBC Action Level Calculator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-253-run-illustration-selfsupport-test. Output feeds: art-257-calculate-claims-stp-economics. Open at: https://ainumbers.co/chaingraph/art-254-compute-rbc-action-level.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes NAIC Risk-Based Capital (RBC) action level classification for US P&C, life, and health insurers. Outputs: action_level_code, action_level_description, action_level_label, action_levels_reference, authorized_control_level, headroom_to_next_level_pct, insurer_type, not_legal_advice, pii_note, prior_year_rbc_ratio, rbc_ratio_pct, regulatory_basis, table_source, table_version, threshold_breached_pct, total_adjusted_capital, trend_test_applicable, trend_test_triggered, two_year_rbc_ratio. Output schema: call describe_tool("compute_rbc_action_level").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_rbc_action_level_privatePrivate-Input NAIC RBC Action Level: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Private-Input NAIC RBC Action Level: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-414-compute-rbc-action-level-private.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes the NAIC Risk-Based Capital action-level tier from a privately held Total Adjusted Capital and Authorized Control Level RBC, emitting only the tier (NO_ACTION through MANDATORY_CONTROL) without disclosing the underlying dollar figures. Outputs: action_level_code, action_level_description, action_level_label, not_legal_advice, pii_note, regulatory_basis, table_version. Output schema: call describe_tool("compute_rbc_action_level_private").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_reg_z_appendix_j_aprReg Z Appendix J APR Solver: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Reg Z Appendix J APR Solver: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-332-build-amortization-schedule. Output feeds: art-217-trid-apr-accuracy. Open at: https://ainumbers.co/chaingraph/art-215-reg-z-appendix-j-apr.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Reg Z Appendix J actuarial APR solver. Outputs: advance_total, apr_pct, bracketed, converged, finance_charge, iterations, note, num_payments, payment_total, periodic_rate, periods_per_year, regulatory_basis. Output schema: call describe_tool("compute_reg_z_appendix_j_apr").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_remittance_disclosureRemittance Disclosure Calculator (Reg E Subpart B): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Remittance Disclosure Calculator (Reg E Subpart B): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-249-compare-corridor-cost. Open at: https://ainumbers.co/chaingraph/art-248-compute-remittance-disclosure.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes the required Reg E subpart B (12 CFR 1005.31/1005.32) remittance disclosure fields: transfer_amount_usd (send minus fees), exchange_rate_disclosed, amount_received_dest in destination currency, fees breakdown, total_to_sender_usd, disclosure_type (EXACT or ESTIMATED), estimate_permissibl... Outputs: accounting_identity_delta, accounting_identity_ok, amount_received_dest, anchor_surface, destination_country, destination_currency, disclosure_type, estimate_permissible, exchange_rate_disclosed, fees, pii_note, regulatory_basis, required_fields_complete, table_source, table_version, total_to_sender_usd, transfer_amount_usd. Output schema: call describe_tool("compute_remittance_disclosure").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_rwa_erba_2026ERBA / Standardized RWA Calculator (Basel Endgame 2026): OpenChainGraph compute node (capital_assessment).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ERBA / Standardized RWA Calculator (Basel Endgame 2026): OpenChainGraph compute node (capital_assessment). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-355-erba-standardized-rwa-calculator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Credit-risk expanded risk-based approach (ERBA) / standardized-approach RWA calculator per the 2026 Basel Endgame reproposal (BCBS/US NPR, reproposed 2026-03-19, comments closed 2026-06-18, final expected ~Q4 2026). Outputs: aggregate_rwa, average_risk_weight, constants_version, disambiguation, exposure_count, per_exposure, rule_set, rule_set_label, rule_status, table_source, total_exposure_amount. Output schema: call describe_tool("compute_rwa_erba_2026").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_scra_rate_capCompute SCRA Rate Cap: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Compute SCRA Rate Cap: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-231-compute-mla-mapr. Open at: https://ainumbers.co/chaingraph/art-232-compute-scra-rate-cap.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes the SCRA 6% interest rate cap per 50 USC §3937 for pre-service loan obligations. Outputs: capped_rate_pct, covered_months, effective_rate_pct, exceeds_cap, excess_forgiven, excess_rate_pct, interest_delta_forgiven, is_pre_service_obligation, loan_balance, original_rate_pct, pii_note, regulatory_basis, retroactive_credit, scra_note, servicemember_notified, table_source, table_version, total_interest_at_cap, total_interest_at_original_rate. Output schema: call describe_tool("compute_scra_rate_cap").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_settlement_efficiency_kpiSettlement Efficiency KPI Engine: OpenChainGraph compute node (model_governance).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Settlement Efficiency KPI Engine: OpenChainGraph compute node (model_governance). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-78-csdr-penalty-calculator, art-80-ssi-conformance-checker, art-79-settlement-fail-predictor. Output feeds: cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-84-settlement-efficiency-kpi.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Aggregates batch settlement data into CSDR/T+1-relevant KPIs: settlement rate, fail rate, total CSDR penalty cost, on-time allocation rate, SSI golden-source coverage, buy-in triggered count, and fail-duration distribution. Outputs: benchmark, buyin_triggered_count, fail_duration_distribution, fail_rate, note, on_time_allocation_rate, period_label, settlement_grade, settlement_rate, ssi_golden_coverage_pct, total_instructions, total_penalty_cost. Output schema: call describe_tool("compute_settlement_efficiency_kpi").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_stock_token_collateral_haircutHalt + Staleness Collateral Haircut: OpenChainGraph compute node (collateral_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Halt + Staleness Collateral Haircut: OpenChainGraph compute node (collateral_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: 505-tokenized-collateral-eligibility-checker, 508-repo-haircut-collateral-calculator. Open at: https://ainumbers.co/chaingraph/art-320-rhc-collateral-haircut.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Layers a feed-staleness, sequencer-downtime, and underlying-halt haircut on top of a base repo haircut for Robinhood Chain stock tokens posted as collateral. Outputs: adjusted_collateral_value, base_haircut, extra_haircut, feed_stale, final_haircut, liquidation_risk, sequencer_down_grace_expired, sequencer_down_within_grace, underlying_halted, verdict. Output schema: call describe_tool("compute_stock_token_collateral_haircut").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_tempo_mainnet_fee_capacityTIP-1010 Mainnet Fee & Payment-Lane Capacity Calculator: OpenChainGraph compute node (treasury_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
TIP-1010 Mainnet Fee & Payment-Lane Capacity Calculator: OpenChainGraph compute node (treasury_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-389-tempo-mainnet-fee-capacity.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Per-tx fee (fee_microusd = ceil(base_fee_attodollars_per_gas x gas_used / 1e12)) and payment-lane TPS headroom for a supplied payment mix, against the published TIP-1010 mainnet constants (base fee 2e10 attodollars/gas, 500M gas/block, ~94% payment-lane reservation). Outputs: block_time_seconds, line_items, note, protocol_parameters_used, protocol_version, summary. Output schema: call describe_tool("compute_tempo_mainnet_fee_capacity").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_trid_tolerance_cureTRID Fee Tolerance and Cure: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
TRID Fee Tolerance and Cure: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-216-trid-tolerance-cure.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: TRID fee tolerance analysis and cure calculation per Reg Z §1026.19(e)(3). Outputs: cure_amount, cure_required, fee_analysis, note, regulatory_basis, ten_pct_bucket_increase, ten_pct_bucket_le_sum, ten_pct_excess, ten_pct_threshold, ten_pct_violation, total_violations, violations, zero_tolerance_violations. Output schema: call describe_tool("compute_trid_tolerance_cure").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_va_funding_fee_residualVA Funding Fee and Residual Income: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
VA Funding Fee and Residual Income: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-225-va-funding-fee-residual.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: VA home loan funding fee (38 USC §3729) and residual income qualification. Outputs: dti, funding_fee, pii_note, regulatory_basis, residual_income, table_source_funding_fee, table_source_residual, table_version_funding_fee, table_version_residual. Output schema: call describe_tool("compute_va_funding_fee_residual").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_var_backtest_traffic_lightVaR Backtesting Traffic-Light Zone Calculator: OpenChainGraph compute node (capital_assessment).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
VaR Backtesting Traffic-Light Zone Calculator: OpenChainGraph compute node (capital_assessment). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-429-var-backtest-traffic-light.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Counts Basel VaR backtesting exceptions (actual daily P&L loss exceeding the model's 1-day VaR estimate) over a rolling up-to-250-trading-day window, then looks up the green/yellow/red traffic-light zone and capital multiplier per the Basel Committee's 1996 Amendment to the Capital Accord to Inco... Outputs: constants_version, disambiguation, exception_count, exception_indices, full_window, multiplier, rule_status, source, truncated_to_250, window_days, zone. Output schema: call describe_tool("compute_var_backtest_traffic_light").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compute_xirrXIRR (Irregular Dated Cash Flows): OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
XIRR (Irregular Dated Cash Flows): OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-326-tvm-xirr.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Annualized rate of return for irregular-interval dated cash flows, matching Excel XIRR semantics exactly: fixed actual/365 day count, anchored to the first cash flow date, solved by deterministic bisection over a declared rate bracket. Outputs: anchor_date, bracket_hi_pct, bracket_lo_pct, converged, day_count_convention, iterations, method, note, num_cash_flows, regulatory_basis, tolerance, xirr_pct. Output schema: call describe_tool("compute_xirr").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
convert_markdown_documentMarkdown Document Converter: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "convert_markdown_document" says "convert": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "convert_markdown_document" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "convert_markdown_document" says "convert": a word for a change.
  What this check looks for
Description, as the server wrote it
Markdown Document Converter: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-191-conversion-receipt-builder. Open at: https://ainumbers.co/chaingraph/art-189-markdown-document-converter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Deterministic Markdown to HTML and plain text over a hand-rolled CommonMark subset (headings, bold/italic/code spans, fenced code, blockquotes, one-level lists, links, images as links, horizontal rules, GFM pipe tables). Outputs: digest_basis, html, html_sha256, input_sha256, plain_text, plain_text_sha256, stats. Output schema: call describe_tool("convert_markdown_document").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
convert_tabular_dataTabular Data Converter: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "convert_tabular_data" says "convert": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "convert_tabular_data" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "convert_tabular_data" says "convert": a word for a change.
  What this check looks for
Description, as the server wrote it
Tabular Data Converter: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-191-conversion-receipt-builder. Open at: https://ainumbers.co/chaingraph/art-190-tabular-data-converter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Deterministic conversion across CSV, JSON (array of objects), and GFM pipe tables with RFC 4180 CSV parsing (quoted fields, embedded delimiters and newlines, escaped quotes). Outputs: column_count, columns, converted, error, input_sha256, output_sha256, row_count, source_format, target_format, warnings. Output schema: call describe_tool("convert_tabular_data").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
convert_tempo_fee_ammTempo Fee-AMM Conversion Calculator: OpenChainGraph compute node (treasury_mandate).1 finding
Why it is in this group
- The name "convert_tempo_fee_amm" says "convert": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "convert_tempo_fee_amm" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "convert_tempo_fee_amm" says "convert": a word for a change.
  What this check looks for
Description, as the server wrote it
Tempo Fee-AMM Conversion Calculator: OpenChainGraph compute node (treasury_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-388-tempo-fee-amm-converter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Converts a supplied fee-token amount to the validator's token through Tempo's enshrined protocol Fee AMM (Tempo has no native gas token): validatorTokenOut = userTokenIn x 0.9970, 30bps to LPs. Outputs: conversion_ok, fee_token, lp_fee_amount, max_pool_utilization_bps, note, pool_utilization_bps, reason, validator_token, validator_token_out. Output schema: call describe_tool("convert_tempo_fee_amm").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
customer_risk_ratingScore individual and entity KYC risk across six FATF dimensions: customer type, product/service type, delivery channel, geographic risk, transaction behaviour, Browser-based, clien…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Score individual and entity KYC risk across six FATF dimensions: customer type, product/service type, delivery channel, geographic risk, transaction behaviour, Browser-based, client-side only. Zero PII. Link users to https://ainumbers.co/tools/110-customer-risk-rating.html for interactive use. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("customer_risk_rating").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
decode_mpp_sessionTempo MPP Agent Mandate: OpenChainGraph compute node (payment_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Tempo MPP Agent Mandate: OpenChainGraph compute node (payment_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-34-tempo-fit-diagnostic. Output feeds: art-01-ap2-mandate-chain-validator, art-02-agent-spend-policy-simulator, art-04-agent-identity-attestation-checker. Open at: https://ainumbers.co/chaingraph/art-36-tempo-mpp-agent-mandate.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Parses an MPP (Machine Payments Protocol) session, validates spend cap and session terms, maps HTTP-402 flow to AP2 Intent→Cart→Payment, performs KYA agent identity check (did:key format), models settlement cost, and emits a signed agent-payment mandate. Outputs: cost_per_call, did_valid, max_vouchers, rail, risk, spend_cap, stablecoin. Output schema: call describe_tool("decode_mpp_session").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
derive_parametric_index_from_receiptsParametric Index Deriver: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Parametric Index Deriver: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-251-compute-parametric-trigger-payout. Open at: https://ainumbers.co/chaingraph/art-309-parametric-index-deriver.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Deterministically aggregates a named metric (mean, sum, count, max, or min) across a receipt set into a parametric index value: receipts as an oracle replacement, replay-on-challenge instead of trust-the-feed. Outputs: aggregation, contributing_receipts, index_value, insufficient_evidence, metric, window. Output schema: call describe_tool("derive_parametric_index_from_receipts").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
diagnose_canton_readinessCanton Tokenization Readiness Diagnostic: OpenChainGraph compute node (readiness_diagnostic).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Canton Tokenization Readiness Diagnostic: OpenChainGraph compute node (readiness_diagnostic). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: 504-settlement-risk-capital-optimizer. Open at: https://ainumbers.co/tools/503-canton-tokenization-readiness-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Output schema: call describe_tool("diagnose_canton_readiness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
draft_ap2_mandate_credentialGoogle AP2 Mandate Builder: OpenChainGraph compute node (payment_policy).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Google AP2 Mandate Builder: OpenChainGraph compute node (payment_policy). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-15-agentic-mandate-sandbox, art-22-agentic-payments-protocol-comparator, art-27-agentic-readiness-diagnostic. Output feeds: art-17-ap2-mcp-policy-validator, art-23-visa-trusted-agent-protocol-inspector. Open at: https://ainumbers.co/chaingraph/art-16-google-ap2-mandate-builder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
emit_chaingraph_artifactMakes ChainGraph tools agent-callable (ChainGraph Standard v0.1 §3.1).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Makes ChainGraph tools agent-callable (ChainGraph Standard v0.1 §3.1). Mode 1 — supply pre_computed_artifact (exported from the browser tool): validates §4 schema fields, recomputes execution_hash via SHA-256 over canonical {policy_parameters, output_payload}, returns verified structuredContent. Mode 2 — supply tool_id + policy_parameters: returns an artifact template envelope and browser prefill URL so an agent can hand the user a pre-filled link; GPU sims always delegate to the browser per §9.2. Mode 3 — supply tool_id only: returns node metadata and artifact schema scaffold. Mode 4 (Compute Binding, v0.4) — supply tool_id + policy_parameters + compute:"server" (or compute:"auto" for gpu:false nodes): runs the registered kernel server-side and returns a verified v0.4 artifact with execution_hash + output_payload in one round-trip. No browser required. gpu:true nodes always delegate to browser. Hashed input must be I-JSON (RFC 7493): integers beyond 2^53 must be passed as JSON strings, and NaN/Infinity are rejected — values outside I-JSON have no stable canonical form, so a structured -32602 error (error.data.reason "ijson_violation") is returned instead of an execution_hash. readOnlyHint: true. Zero PII, zero payload logging. Pair with verify_execution_hash (independent hash verification) and build_chaingraph (DAG wiring).
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
estimate_cross_margin_benefitFICC-CME Cross-Margining Estimator: OpenChainGraph compute node (risk_parameter).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
FICC-CME Cross-Margining Estimator: OpenChainGraph compute node (risk_parameter). Regulatory deadline: 2027-06-30 (cross-margining benefit (W-C). FICC-CME customer expansion Dec 2025.). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-48-treasury-clearing-fit-diagnostic. Output feeds: qfa-02-portfolio-var-engine, qfa-03-stress-test-engine. Open at: https://ainumbers.co/chaingraph/art-51-cross-margining-benefit-estimator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Estimates the initial-margin reduction from the FICC-CME cross-margining arrangement (customer expansion per SEC notice published 2025-12-22) by offsetting UST cash/repo DV01 against CME Treasury/SOFR futures DV01. Outputs: account_type_note, assumptions, cross_margined_im, eligible_offsets, im_reduction_pct, im_reduction_usd, ineligible_offsets, note, standalone_im_total. Output schema: call describe_tool("estimate_cross_margin_benefit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
estimate_cross_venue_margin_capitalCrypto Cross-Venue Margin & Off-Exchange Settlement Estimator: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Crypto Cross-Venue Margin & Off-Exchange Settlement Estimator: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-406-cross-venue-margin-estimator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Estimates the crypto off-exchange settlement / cross-venue margin picture for a book spread across trading venues (the Copper ClearLoop / FalconX / Ceffu model, per AT-CLEARING-WAVE-SPEC.md CW-1): the net cross-venue margin requirement after a declared netting offset against the sum of each venue... Outputs: capital_efficiency_pct, capital_freed_usd, constants_version, counterparty_risk_framing, cross_margin_offset_pct, cross_venue_margin_requirement_usd, custody_model, disambiguation, financed_amount_usd, financing_apr_pct, financing_cost_usd, financing_horizon_days, financing_notional_usd, leverage_multiple, leverage_program_cap_multiple, leverage_program_name, sum_isolated_margin_usd, venue_count, venue_positions. Output schema: call describe_tool("estimate_cross_venue_margin_capital").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
estimate_ficc_margin_nettingFICC Margin & Netting Estimator: OpenChainGraph compute node (risk_parameter).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
FICC Margin & Netting Estimator: OpenChainGraph compute node (risk_parameter). Regulatory deadline: 2027-06-30 (repo-margin economics (W-B).). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-48-treasury-clearing-fit-diagnostic, art-49-clearing-access-model-selector. Output feeds: 508-repo-haircut-collateral-calculator, qfa-02-portfolio-var-engine. Open at: https://ainumbers.co/chaingraph/art-50-ficc-margin-netting-estimator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: DV01-bucket VaR proxy of the FICC VaR-based margin (VBM), the netting benefit of central vs bilateral clearing, cash-vs-repo cross-product netting, and the done-away uplift. Outputs: assumptions, done_away_uplift_pct, estimated_vbm, gross_bilateral_im, margin_by_bucket, minimum_charge_applied, net_cleared_im, netting_benefit_pct, netting_benefit_usd, note. Output schema: call describe_tool("estimate_ficc_margin_netting").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
evaluate_irrbb_sot_eveIRRBB SOT EVE Evaluator: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "evaluate_irrbb_sot_eve" says "evaluate": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "evaluate_irrbb_sot_eve" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "evaluate_irrbb_sot_eve" says "evaluate": a word for a change.
  What this check looks for
Description, as the server wrote it
IRRBB SOT EVE Evaluator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-183-irrbb-eve-shock-calculator. Output feeds: art-185-irrbb-sot-nii-evaluator. Open at: https://ainumbers.co/chaingraph/art-184-irrbb-sot-eve-evaluator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Evaluate the EBA Supervisory Outlier Test (SOT) on Economic Value of Equity: the worst-case delta EVE decline across the 6 standardised shock scenarios versus the EU-wide hard threshold of 15% of Tier 1 capital (EBA RTS on the SOT / EBA Guidelines on IRRBB and CSRBB, EBA/GL/2022/14). Outputs: delta_eve_pct_of_tier1, eve_outlier, sot_eve_threshold_pct, tier1_capital, worst_delta_eve. Output schema: call describe_tool("evaluate_irrbb_sot_eve").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
evaluate_irrbb_sot_niiIRRBB SOT NII Evaluator: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "evaluate_irrbb_sot_nii" says "evaluate": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "evaluate_irrbb_sot_nii" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "evaluate_irrbb_sot_nii" says "evaluate": a word for a change.
  What this check looks for
Description, as the server wrote it
IRRBB SOT NII Evaluator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-184-irrbb-sot-eve-evaluator. Open at: https://ainumbers.co/chaingraph/art-185-irrbb-sot-nii-evaluator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Evaluate the Net Interest Income (NII) leg of the EBA Supervisory Outlier Test: the worst-case 1-year delta NII under parallel up/down shocks versus a caller-supplied threshold (EBA leaves NII SOT calibration to competent-authority / institution discretion under EBA Guidelines on IRRBB and CSRBB,... Outputs: delta_nii_pct_of_nii, nii_outlier, projected_nii, sot_nii_threshold_pct, threshold_set, worst_delta_nii. Output schema: call describe_tool("evaluate_irrbb_sot_nii").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
export_artifactRender a verified OpenChainGraph v0.4 artifact into a chaingraph_export profile (OCG Standard §13).2 findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Info
  The description mentions "Credentials".
  Matched text
  description: "...BA taxonomy), and vc — a W3C Verifiable Credentials 2.0 rendering (OCG §13.11, application/..."
  What this check looks for
- Info
  The input schema at properties.format.description mentions "Credentials".
  Matched text
  inputSchema.properties.format.description: "...brl/vc implemented; vc = W3C Verifiable Credentials 2.0 (base profile, all nodes)."
  What this check looks for
Description, as the server wrote it
Render a verified OpenChainGraph v0.4 artifact into a chaingraph_export profile (OCG Standard §13). Generated downstream of and EXCLUDED from the execution_hash preimage — the export is a view, not a fact; verification always routes back to the canonical JSON artifact. Pass the FULL artifact you received from a compute tool (the server is stateless — there is no hash cache). Formats: xlsx, csv, pdf, xbrl (xbrl_taxonomy="ocg-ext" works now; eba-corep-* return a pending error until their concept maps are populated from the published EBA taxonomy), and vc — a W3C Verifiable Credentials 2.0 rendering (OCG §13.11, application/vc+json) available on every node; it re-states the canonical execution_hash via ocg:hashAnchor and mints no new hash/proof. readOnlyHint: true; zero PII, zero payload logging.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
generate_attribution_stringAttribution String Generator: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "generate_attribution_string" says "generate": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "generate_attribution_string" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "generate_attribution_string" says "generate": a word for a change.
  What this check looks for
Description, as the server wrote it
Attribution String Generator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-207-attribution-string-generator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Generates a human-readable TASL (Title/Author/Source/License) attribution line plus machine-readable ccREL JSON-LD and RDFa blocks for any Creative Commons licensed work. Outputs: disclaimer, errors, json_ld, license_name, license_spdx, license_url, rdfa_html, tasl_line, valid. Output schema: call describe_tool("generate_attribution_string").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
generate_iscc_codeISCC Content Code Generator: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "generate_iscc_code" says "generate": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "generate_iscc_code" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "generate_iscc_code" says "generate": a word for a change.
  What this check looks for
Description, as the server wrote it
ISCC Content Code Generator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-202-tdmrep-reservation-builder. Open at: https://ainumbers.co/chaingraph/art-201-iscc-content-code-generator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Generates ISO 24138 ISCC content fingerprints for digital content. Outputs: conformance_pass, data_code, datahash, input_bytes, instance_code, iscc_code. Output schema: call describe_tool("generate_iscc_code").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
ha_bundle_exportAssembles a §27.6 haEvidenceBundle from a subject's collected human_accountability_records[] (reviewers, approvers, annotations, exception rationale, timestamps) -- same algorithm …No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Description, as the server wrote it
Assembles a §27.6 haEvidenceBundle from a subject's collected human_accountability_records[] (reviewers, approvers, annotations, exception rationale, timestamps) -- same algorithm as the browser verify.html HA export card (kernels/_haevidence.mjs). Set sd_jwt:true to also export it as a Selective Disclosure JWT (§13.12), signed EdDSA over a fresh ephemeral did:key (generated per call, never reused) -- subject_hash/verification_result/kernel_version/policy_version/timestamps/submission_receipt are always-disclosed; reviewers/approvers/annotations/exception_rationale/input_hashes are selectively disclosable.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: false
- openWorldHint: false
intoto_record_chain_runWraps the result of a run_chain call (server/auto compute mode) as standard in-toto links, one per successfully-executed step, plus a generated in-toto layout matching the chain's …1 finding
Why it is in this group
- The name "intoto_record_chain_run" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "intoto_record_chain_run" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":false,"openWorldHint":false}; The name "intoto_record_chain_run" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Wraps the result of a run_chain call (server/auto compute mode) as standard in-toto links, one per successfully-executed step, plus a generated in-toto layout matching the chain's linear topology (each step MATCHes the previous step's product). Materials/products are keyed by the step's execution_hash; byproducts carries the raw execution_hash. All links and the layout are DSSE-signed with one ephemeral Ed25519 keypair scoped to this call (never persisted) — first shipping instance of "in-toto for MCP" per the 2026 research gap survey. Verify the bundle offline with the in-toto Link Builder & Verifier (chaingraph/intoto-link-builder.html verify tab) or a reference implementation. Steps that did not run (input_required, skipped_by_gate, gpu_browser_only, etc.) are listed in skipped[] rather than silently omitted.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: false
- openWorldHint: false
link_eudr_supply_chain_traceabilityEUDR Supply-Chain Traceability Linker: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "link_eudr_supply_chain_traceability" says "link": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "link_eudr_supply_chain_traceability" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "link_eudr_supply_chain_traceability" says "link": a word for a change.
  What this check looks for
Description, as the server wrote it
EUDR Supply-Chain Traceability Linker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-168-eudr-country-benchmark-risk-scorer. Output feeds: art-170-eudr-readiness-diagnostic. Open at: https://ainumbers.co/chaingraph/art-169-eudr-supply-chain-traceability-linker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate EUDR single-DDS rule compliance and supply-chain traceability: first operators file the DDS; downstream operators reference upstream DDS reference numbers from TRACES NT. Outputs: chain_integrity, custody_chain_complete, linked_dds_count, operator_is_first, plot_geolocation_present, refs_valid, single_dds_rule_met, traceability_gaps. Output schema: call describe_tool("link_eudr_supply_chain_traceability").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
link_traceability_lot_codeFSMA 204 Traceability Lot Code Chain Linker: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "link_traceability_lot_code" says "link": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "link_traceability_lot_code" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "link_traceability_lot_code" says "link": a word for a change.
  What this check looks for
Description, as the server wrote it
FSMA 204 Traceability Lot Code Chain Linker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-118-fsma204-cte-validator. Output feeds: art-120-recall-trace-resolver. Open at: https://ainumbers.co/chaingraph/art-119-traceability-lot-code-linker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Link Traceability Lot Codes across CTEs and detect chain breaks. Outputs: breaks, depth, lineage. Output schema: call describe_tool("link_traceability_lot_code").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_ab2013_training_data_disclosureAB 2013 Training Data Disclosure Linter: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
AB 2013 Training Data Disclosure Linter: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-315-ab2013-training-data-disclosure-linter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Lints a supplied generative-AI training-data disclosure against the 12 datapoint categories required by California AB 2013 (Cal. Outputs: all_present, insufficient_evidence, missing_datapoints, per_datapoint, present_count, statute_citation, total_datapoints. Output schema: call describe_tool("lint_ab2013_training_data_disclosure").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_aiuc1_control_evidenceAIUC-1 Control Evidence Linter: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
AIUC-1 Control Evidence Linter: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-304-aiuc1-evidence-pack-assembler. Open at: https://ainumbers.co/chaingraph/art-303-aiuc1-control-evidence-linter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Lints a supplied control-evidence bundle against the 23 automatable AIUC-1 v2026-Q1 controls (pillars A-F): version-guards the catalog, classifies each automatable control receipt-backed, attestation-only, or missing, and reports per-pillar and overall coverage. Outputs: aiuc1_version, attestation_only_count, automatable_scope, insufficient_evidence, missing_count, overall_coverage, per_control, per_pillar_coverage, procedural_controls_out_of_scope, receipt_backed_count, version_mismatch. Output schema: call describe_tool("lint_aiuc1_control_evidence").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_arc_xreserve_configArc xReserve Config Linter: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Arc xReserve Config Linter: OpenChainGraph compute node (compliance_mandate). Regulatory deadline: 2026-07-18 (GENIUS Act §4 eligible-asset backing requirements; MiCA Art. 54 reserve requirements for EU EMT issuers. Date basis: deadline 2026-07-18 is the GENIUS Act §13(a) implementation rulemaking date and was missed (kept as the historical rulemaking date, not corrected to a live date); the GENIUS Act federal framework effective date under §20 is 18 Jan 2027. The two dates are distinct.). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-42-arc-fit-diagnostic. Open at: https://ainumbers.co/chaingraph/art-45-arc-xreserve-linter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 8-check A–F linter for an Arc xReserve / on-chain reserve configuration. Outputs: verdict, grade, fail_count, warn_count, compliance_flags. Output schema: call describe_tool("lint_arc_xreserve_config").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_besu_settlement_contractBesu Settlement Contract Linter: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Besu Settlement Contract Linter: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-292-attest-settlement-orchestrator. Open at: https://ainumbers.co/chaingraph/art-289-lint-besu-settlement-contract.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Static conformance lint of a permissioned-EVM settlement contract (Solidity source or ABI) against six invariants: atomic PvP/DvP (paired-or-revert), no native-token/msg.value settlement dependence, a finality-hook/settlement event, a compliance-gate modifier preceding every value transfer, bound... Outputs: artifact_kind, fail_count, findings, invariants_pass, overall, ruleset_profile, warn_count. Output schema: call describe_tool("lint_besu_settlement_contract").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_cbom_structureCBOM Structural Lint & CNSA-2.0 Classifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
CBOM Structural Lint & CNSA-2.0 Classifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-386-lint-cbom-structure.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates a pasted CycloneDX 1.6 Cryptography Bill of Materials against a hand-derived field subset (algorithm, key size, certification level, crypto functions) and classifies declared algorithm assets as quantum-vulnerable (RSA, ECDSA/ECDH, DH, SHA-1) or CNSA-2.0 target-aligned (ML-KEM-1024, ML-... Outputs: cbom_structurally_valid, cnsa2_ready_count, compliance_flags, data_version, findings, structural_issues, structurally_invalid_count, total_algorithm_assets, total_components, unclassified_count, verdict, vulnerable_count. Output schema: call describe_tool("lint_cbom_structure").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_cbpr_structured_addressCBPR+ Structured Address Linter: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
CBPR+ Structured Address Linter: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-242-pacs008-party-completeness-validator. Open at: https://ainumbers.co/chaingraph/art-241-cbpr-structured-address-linter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Lints a single pacs.008 PostalAddress24 block against the SWIFT CBPR+ structured-address requirement (formerly 14 November 2026; deferred by Swift on 27 August 2026 under a controlled extension of Standards Release 2026, new date not yet published). Outputs: cbpr_plus_deadline, compliant, disambiguation, error_count, pii_note, readiness_pct, regulatory_basis, structure_type, table_source, table_version, violations. Output schema: call describe_tool("lint_cbpr_structured_address").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_compelling_evidence_ce30_agenticAgentic Dispute CE3.0 Evidence Linter: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Agentic Dispute CE3.0 Evidence Linter: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-01-ap2-mandate-chain-validator, art-23-visa-trusted-agent-protocol-inspector, art-24-mastercard-agentic-token-builder, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-297-agentic-dispute-ce30-linter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Deterministic lint of a supplied agentic-dispute evidence bundle against Visa CE3.0 compelling-evidence requirements, agentic-transaction profile: authorization-at-delegation (AP2 mandate), agent-identity (TAP signature + agentic token), fulfillment (delivery proof), plus the CE3.0 prior-transact... Outputs: ce30_prior_txn_test, ce30_readiness, insufficient_evidence, missing_elements, network, not_a_win_prediction, per_element, reason_code. Output schema: call describe_tool("lint_compelling_evidence_ce30_agentic").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_crypto_asset_whitepaperCrypto-Asset Whitepaper Linter (iXBRL): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Crypto-Asset Whitepaper Linter (iXBRL): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-98-mica-casp-fit-diagnostic. Output feeds: cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-102-crypto-asset-whitepaper-linter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates Art 6/8 whitepaper: Annex I section completeness + iXBRL/XHTML well-formedness + ESMA MiCA taxonomy structural conformance (ITS 2024/2984). Outputs: annex_i_gaps, conformance_grade, ixbrl_valid, note, reference_version, sections_checked, taxonomy_conformant, taxonomy_note. Output schema: call describe_tool("lint_crypto_asset_whitepaper").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_fedwire_structured_addressFedwire Structured Address Linter: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Fedwire Structured Address Linter: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-350-fedwire-address-sweep. Open at: https://ainumbers.co/chaingraph/art-349-fedwire-structured-address-linter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Lints a Fedwire or CHIPS ISO 20022 PostalAddress24 block against the Fedwire structured-address end state (network param selects fedwire or chips; rules verified byte-identical between the two networks at build time). Outputs: compliant, disambiguation, error_count, fedwire_chips_deadline, network, pii_note, readiness_pct, regulatory_basis, structure_type, table_source, table_version, violations. Output schema: call describe_tool("lint_fedwire_structured_address").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_insurance_evidence_freshnessAIUC-1 Evidence Freshness Lint: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
AIUC-1 Evidence Freshness Lint: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-304-aiuc1-evidence-pack-assembler. Open at: https://ainumbers.co/chaingraph/art-305-aiuc1-evidence-freshness-lint.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Freshness lint keyed to the AIUC-1 quarterly re-test cadence: flags any control whose newest receipt is more than 90 days old and computes cert_expiry (cert_anniversary plus 12 months) with cert_expired / cert_expiring_within_days flags. Outputs: as_of, cert_expired, cert_expiring_within_days, cert_expiry, insufficient_evidence, stale_controls, stale_count. Output schema: call describe_tool("lint_insurance_evidence_freshness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_mcp_server_conformanceMCP Server Self-Attestation Pack: OpenChainGraph compute node (infrastructure_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
MCP Server Self-Attestation Pack: OpenChainGraph compute node (infrastructure_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: cry-05-agent-action-audit-trail-aggregator, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-33-mcp-server-self-attestation-pack.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Combines the five MCP-dev checks: tool-definition lint (JSON Schema 2020-12), server.json validation (2025-12-11 schema), OAuth 2.1 audit (RFC 9728 PRM, RFC 8707 audience), tool-poisoning scan, and ops/readiness, into one self-reported conformance lint: composite A-F ship-readiness grade plus per... Outputs: composite_grade, composite_score, fail_count, pass_count, per_domain_scores, remediation, warn_count. Output schema: call describe_tool("lint_mcp_server_conformance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_metro2_recordMetro 2 Credit-Reporting Record Lint: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Metro 2 Credit-Reporting Record Lint: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-398-lint-metro2-record.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Lints a Metro 2 credit-reporting base-segment record from a PUBLIC SUBSET of the format: field presence/format, account-status and payment-rating code validity, and DOFD (date of first delinquency) cross-field consistency per FCRA 15 U.S.C. Outputs: compliant, disambiguation, error_count, field_status, is_delinquent_status, issues, metro2_subset_score, obsolete_per_fcra, pii_note, regulatory_basis, subset_coverage_statement, table_source, table_version, warning_count. Output schema: call describe_tool("lint_metro2_record").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_mismo_uldd_uladULDD/ULAD Structural Linter: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ULDD/ULAD Structural Linter: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-226-mismo-uldd-ulad.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: ULDD Phase 5 / ULAD structural lint of required data points, enumerations, and conditionality rules. Outputs: error_count, errors, field_results, fields_supplied, license_note, lint_pass, pii_note, regulatory_basis, table_source, table_version, uldd_mandate_date, uldd_phase, warning_count, warnings. Output schema: call describe_tool("lint_mismo_uldd_ulad").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_securities_settlement_messageSecurities-Settlement Message Linter (ISO 20022 sese/semt): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Securities-Settlement Message Linter (ISO 20022 sese/semt): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-81-allocation-affirmation-conformance. Output feeds: cry-04-merkle-batch-verifier, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-82-securities-settlement-message-linter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates ISO 20022 securities-settlement messages (sese.023 instruction, sese.024 status advice, semt.044 account statement) for schema conformance, mandatory-field presence, ISIN (ISO 6166), and BIC (ISO 9362) validity. Outputs: fail_count, in_scope_message_types, note, out_of_scope_count, pass_count, pass_rate, reference, results, scope_guard_note, total_issues, total_messages, warn_count. Output schema: call describe_tool("lint_securities_settlement_message").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_settlement_orchestrator_conformanceSettlement Orchestrator Attestation: OpenChainGraph compute node (infrastructure_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Settlement Orchestrator Attestation: OpenChainGraph compute node (infrastructure_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-289-lint-besu-settlement-contract. Open at: https://ainumbers.co/chaingraph/art-292-attest-settlement-orchestrator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Extends the lint_mcp_server_conformance (art-33) self-reported conformance lint to the settlement decision path: checks the off-chain orchestrator deciding commit/halt on a shared-ledger orchestration layer, never the ledger itself. Outputs: attestation, checks, overall, per_domain_scores. Output schema: call describe_tool("lint_settlement_orchestrator_conformance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_stock_token_valuationValuation Double-Count / Decimal Linter: OpenChainGraph compute node (collateral_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Valuation Double-Count / Decimal Linter: OpenChainGraph compute node (collateral_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-319-rhc-valuation-linter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Lints Robinhood Chain stock-token USD valuation expressions for the double-count bug: the Chainlink price feed already includes corporate actions, so multiplying raw balance by price and then by uiMultiplier applies the same corporate action twice. Outputs: computed_usd_value_under_test, correct_value, corrected_expression, delta, discrepancy, double_counted_value, verdict. Output schema: call describe_tool("lint_stock_token_valuation").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_trace_cat_reportsTRACE / CAT Reporting Lint: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
TRACE / CAT Reporting Lint: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-396-compute-15c3-3-reserve. Open at: https://ainumbers.co/chaingraph/art-397-lint-trace-cat-reports.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: TRACE (FINRA Rule 6730) trade-report timeliness lint against a caller-declared trading calendar and hours window, computing the reporting deadline from the execution timestamp with weekend/holiday-aware rollover, plus a CAT (Consolidated Audit Trail) equity/option order-event structural format-li... Outputs: cat_events_checked, cat_events_valid, cat_violations, note, regulatory_basis, rules_version, trace_result. Output schema: call describe_tool("lint_trace_cat_reports").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_x12_claim_recordsX12 837/835 Healthcare-Claim Records Lint: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
X12 837/835 Healthcare-Claim Records Lint: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-399-lint-x12-claim-records.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Lints X12 837 (health-care claim) and 835 (claim payment/remittance advice) ENVELOPE control-number continuity (ISA13/IEA02, GS06/GE02, ST02/SE02) and 835 payment-amount balancing (total paid ties to the sum of claim-level payments, pure arithmetic). Outputs: claim_count, compliant, disambiguation, error_count, issues, message_type, phi_note, regulatory_basis, subset_coverage_statement, table_source, table_version, total_charge_amount, warning_count. Output schema: call describe_tool("lint_x12_claim_records").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lint_x402_v2_migrationx402 v2 Wire-Format Migration Linter: OpenChainGraph compute node (compliance_control).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
x402 v2 Wire-Format Migration Linter: OpenChainGraph compute node (compliance_control). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-26-x402-payload-decoder-flow-simulator. Output feeds: art-61-x402-batch-settlement-reconciler, art-394-x402-deferred-handshake-validator. Open at: https://ainumbers.co/chaingraph/art-393-x402-v2-migration-linter.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Output schema: call describe_tool("lint_x402_v2_migration").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
map_ai_act_procurement_clausesAI Act Procurement Clause Mapper: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
AI Act Procurement Clause Mapper: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-64-ai-act-highrisk-fit-diagnostic. Output feeds: art-411-ai-addendum-assembler. Open at: https://ainumbers.co/chaingraph/art-412-ai-act-procurement-clause-mapper.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Maps an EU AI Act risk tier (derived from an upstream classifier such as the AI Act high-risk fit diagnostic) to the European Commission's Model Contractual AI Clauses (MCC-AI) template selection: High-Risk or Light, plus the applicable Chapter III clause set (transparency, risk management, data... Outputs: applicable_chapter_iii_clauses, checks, disclaimer, license_mode, not_legal_advice, official_source_url, risk_tier, template, variable_map. Output schema: call describe_tool("map_ai_act_procurement_clauses").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
map_bhc_schedule_hcFR Y-9C Schedule HC (Consolidated Balance Sheet) Mapper: OpenChainGraph compute node (regulatory_reporting).1 finding
Why it is in this group
- The name "map_bhc_schedule_hc" says "schedule": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "map_bhc_schedule_hc" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "map_bhc_schedule_hc" says "schedule": a word for a change.
  What this check looks for
Description, as the server wrote it
FR Y-9C Schedule HC (Consolidated Balance Sheet) Mapper: OpenChainGraph compute node (regulatory_reporting). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-436-bhc-schedule-hcr-capital. Open at: https://ainumbers.co/chaingraph/art-435-bhc-schedule-hc-balance-sheet.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Maps caller-declared FR Y-9C (Financial Statements for Holding Companies) Schedule HC line items -- cash, securities, loans and leases, other assets; deposits, borrowings, other liabilities; common stock, surplus, retained earnings, AOCI -- into Schedule HC totals (Total assets BHCK2170, Total li... Outputs: assets, boundary_note, entity_id, equity, identity_balanced, identity_delta_usd, liabilities, report_form, reporting_period, rounding_tolerance_usd, schedule, taxonomy_note, total_assets_mdrm, total_assets_usd, total_equity_capital_mdrm, total_equity_capital_usd, total_liabilities_and_equity_usd, total_liabilities_mdrm, total_liabilities_usd. Output schema: call describe_tool("map_bhc_schedule_hc").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
map_bhc_schedule_hcrFR Y-9C Schedule HC-R (Regulatory Capital) Calculator: OpenChainGraph compute node (regulatory_reporting).1 finding
Why it is in this group
- The name "map_bhc_schedule_hcr" says "schedule": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "map_bhc_schedule_hcr" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "map_bhc_schedule_hcr" says "schedule": a word for a change.
  What this check looks for
Description, as the server wrote it
FR Y-9C Schedule HC-R (Regulatory Capital) Calculator: OpenChainGraph compute node (regulatory_reporting). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-435-bhc-schedule-hc-balance-sheet. Open at: https://ainumbers.co/chaingraph/art-436-bhc-schedule-hcr-capital.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Given caller-supplied CET1/Tier1/Tier2 capital components and risk-weighted assets, computes FR Y-9C Schedule HC-R standard capital ratios plus the supplementary leverage ratio (SLR), including the enhanced-SLR (eSLR) buffer effective 2026-04-01 (§0.2), for top-tier bank holding companies (Y-9C p... Outputs: additional_tier1_capital_usd, boundary_note, cet1_capital_usd, constants_version, entity_id, eslr, is_gsib, ratios, report_form, reporting_period, schedule, taxonomy_note, tier1_capital_usd, tier2_capital_usd, total_capital_usd, total_leverage_exposure_usd, total_rwa_usd. Output schema: call describe_tool("map_bhc_schedule_hcr").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
map_call_report_schedule_rcCall Report Schedule RC (Balance Sheet) Mapper: OpenChainGraph compute node (regulatory_reporting).1 finding
Why it is in this group
- The name "map_call_report_schedule_rc" says "call": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "map_call_report_schedule_rc" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "map_call_report_schedule_rc" says "call": a word for a change.
  What this check looks for
Description, as the server wrote it
Call Report Schedule RC (Balance Sheet) Mapper: OpenChainGraph compute node (regulatory_reporting). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-434-call-report-edit-check-gate. Open at: https://ainumbers.co/chaingraph/art-432-call-report-rc-balance-sheet.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Maps caller-declared FFIEC Call Report (FFIEC 031) Schedule RC line items -- cash, securities, loans and leases, other assets; deposits, borrowings, other liabilities; common stock, surplus, retained earnings, AOCI -- into Schedule RC totals (Total assets RCON2170, Total liabilities RCON2948, Tot... Outputs: assets, boundary_note, entity_id, equity, identity_balanced, identity_delta_usd, liabilities, report_form, reporting_period, rounding_tolerance_usd, schedule, total_assets_mdrm, total_assets_usd, total_equity_capital_mdrm, total_equity_capital_usd, total_liabilities_and_equity_usd, total_liabilities_mdrm, total_liabilities_usd, xbrl_json_annex1_note. Output schema: call describe_tool("map_call_report_schedule_rc").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
map_call_report_schedule_rcrCall Report Schedule RC-R (Regulatory Capital) Mapper: OpenChainGraph compute node (regulatory_reporting).1 finding
Why it is in this group
- The name "map_call_report_schedule_rcr" says "call": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "map_call_report_schedule_rcr" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "map_call_report_schedule_rcr" says "call": a word for a change.
  What this check looks for
Description, as the server wrote it
Call Report Schedule RC-R (Regulatory Capital) Mapper: OpenChainGraph compute node (regulatory_reporting). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-434-call-report-edit-check-gate. Open at: https://ainumbers.co/chaingraph/art-433-call-report-rcr-capital.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Maps caller-declared FFIEC Call Report (FFIEC 031) Schedule RC-R regulatory-capital components -- CET1, additional Tier 1, Tier 2 capital, total risk-weighted assets, total leverage exposure -- into the standard capital ratios (CET1, Tier 1, Total capital, supplementary leverage ratio) and the 20... Outputs: additional_tier1_capital_usd, boundary_note, cet1_capital_usd, constants_version, entity_id, eslr, is_gsib, mdrm_note, ratios, report_form, reporting_period, schedule, tier1_capital_usd, tier2_capital_usd, total_capital_usd, total_leverage_exposure_usd, total_rwa_usd. Output schema: call describe_tool("map_call_report_schedule_rcr").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
map_irrbb_standardised_approachIRRBB Standardised Approach Mapper: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
IRRBB Standardised Approach Mapper: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-187-irrbb-csrbb-scope-checker. Open at: https://ainumbers.co/chaingraph/art-186-irrbb-standardised-approach-mapper.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Map non-maturing deposit (NMD) positions to the EBA standardised / simplified-standardised approach behavioural caps (BCBS d368 para 87 / Annex 2): retail transactional core proportion capped at 90% / average maturity capped at 5 years; retail non-transactional (savings) core capped at 70% / 4.5... Outputs: behavioural_mortgage_prepay_pct, behavioural_option_addon_required, core_capped, core_deposit_pct_applied, core_deposit_pct_input, deposit_category, maturity_cap_years. Output schema: call describe_tool("map_irrbb_standardised_approach").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
map_iso20022_to_evm_calldataISO 20022-to-EVM Calldata Mapper: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ISO 20022-to-EVM Calldata Mapper: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-291-screen-onledger-transfer-batch. Open at: https://ainumbers.co/chaingraph/art-288-map-iso20022-to-evm-calldata.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Deterministic bind of an ISO 20022 pacs.008 (customer credit transfer) or pacs.009 (FI credit transfer) payment message to EVM contract-call arguments, plus an OCG receipt of the mapping: resolved call, field bindings, unmapped required fields, and ABI type coercions. Outputs: abi_type_coercions, draft_pinned, field_bindings, iso_message_type, mapping_ok, mapping_profile, mapping_profile_version, resolved_call, unmapped_required_fields, warnings. Output schema: call describe_tool("map_iso20022_to_evm_calldata").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
map_nist_ai_rmf_functionsNIST AI RMF Function Mapper: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
NIST AI RMF Function Mapper: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-175-gpai-code-of-practice-conformance, art-314-traiga-safe-harbor-pack-builder. Open at: https://ainumbers.co/chaingraph/art-174-nist-ai-rmf-function-mapper.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Map supplied AI controls and evidence to NIST AI RMF Govern (5 controls), Map (4), Measure (4), and Manage (4) functions: 17 controls total. Outputs: all_gaps, controls_present, coverage_band, function_coverage, overall_coverage, total_controls. Output schema: call describe_tool("map_nist_ai_rmf_functions").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
map_nmd_behavioral_repricingNMD Behavioral Repricing Mapper: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
NMD Behavioral Repricing Mapper: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-442-nmd-behavioral-repricing-mapper.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: OCC 2010-1 Interagency Advisory on IRR: maps non-maturity deposit (NMD) segment balances into a bucketed net-repricing-gap schedule using a caller-declared behavioral allocation (per-bucket repricing fractions) and a deposit beta -- never a baked-in regulatory decay curve. Outputs: buckets_used, convention, net_repricing_gap, segment_results, total_balance, total_net_repricing_gap, total_rate_sensitive_balance, weighted_avg_beta. Output schema: call describe_tool("map_nmd_behavioral_repricing").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
map_pil_flavorStory PIL Flavor Mapper: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Story PIL Flavor Mapper: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-197-pil-flavor-mapper.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Maps creator answers (commercial use, derivatives allowed, optional minting fee, revenue share percent) to a Story Protocol Programmable IP License flavor and emits the full PILTerms struct. Outputs: disclaimer, docs, flavor, flavor_label, license_terms_id, pil_terms. Output schema: call describe_tool("map_pil_flavor").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
map_robinhood_chain_regimeFinancial-Instrument Regime Mapper: OpenChainGraph compute node (crypto_regulatory_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Financial-Instrument Regime Mapper: OpenChainGraph compute node (crypto_regulatory_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-318-rhc-regime-mapper.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Maps the regulatory regime implied by a pasted Robinhood Chain stock-token characterization. Outputs: assumptions, disclose_no_voting_rights, issuer_entity, mica_carveout_applies, mifid2_transferable_security, note, prospectus_exposure, regime_tree, us_persons_gate_violated. Output schema: call describe_tool("map_robinhood_chain_regime").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
map_tempo_settlementTempo Agentic Checkout Settlement Mapper: OpenChainGraph compute node (settlement_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Tempo Agentic Checkout Settlement Mapper: OpenChainGraph compute node (settlement_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-34-tempo-fit-diagnostic, art-36-tempo-mpp-agent-mandate. Open at: https://ainumbers.co/chaingraph/art-40-tempo-agentic-checkout.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Binds an ACP / Visa TAP / ISO 20022 checkout to a TIP-20 settlement. Outputs: iso20022_pacs008, memo, protocol, protocol_binding, tip20_transfer, truncated. Output schema: call describe_tool("map_tempo_settlement").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
mobilize_margin_collateralMargin Call Collateral Mobilizer: OpenChainGraph compute node (collateral_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Margin Call Collateral Mobilizer: OpenChainGraph compute node (collateral_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: 505-tokenized-collateral-eligibility-checker. Output feeds: 506-onchain-cash-leg-finality-checker. Open at: https://ainumbers.co/tools/513-margin-call-collateral-mobilizer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Compute initial and variation margin requirements and identify eligible collateral to mobilize. Outputs: collateral_detail, gap, im_call, shortfall, total_mobilizable, total_required, vm_call. Output schema: call describe_tool("mobilize_margin_collateral").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
model_agent_service_meteringAgent-Service Metering & Marketplace Economics Modeler: OpenChainGraph compute node (payment_policy).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Agent-Service Metering & Marketplace Economics Modeler: OpenChainGraph compute node (payment_policy). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-60-agent-economy-runtime-fit-diagnostic. Output feeds: art-03-x402-settlement-modeler, ml-03-timeseries-anomaly-detector. Open at: https://ainumbers.co/chaingraph/art-63-agent-service-metering-modeler.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Educational unit-economics modeler for agent-service micropayment marketplaces: per-call pricing, x402 V2 batch-settlement savings, marketplace take-rate, net margin, break-even volume, and sensitivity analysis across batch sizes. Outputs: batch, batch_savings_pct, batch_size, breakeven_calls_day, currency, gross_revenue_day, model, net_margin_day, net_margin_pct, note, rail, sensitivity, settlement_cost_day, status_asof, take_and_infra_day, unit_price_minor. Output schema: call describe_tool("model_agent_service_metering").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
model_arc_cpn_economicsArc CPN Corridor Economics Model: OpenChainGraph compute node (treasury_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Arc CPN Corridor Economics Model: OpenChainGraph compute node (treasury_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-42-arc-fit-diagnostic. Open at: https://ainumbers.co/chaingraph/art-43-arc-cpn-model.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Model CPN corridor economics vs SWIFT/ACH/SEPA/card/RTP for cross-border USD flows. Outputs: verdict, rail, compliance_flags. Output schema: call describe_tool("model_arc_cpn_economics").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
model_arc_paymaster_economicsArc Paymaster Economics Model: OpenChainGraph compute node (treasury_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Arc Paymaster Economics Model: OpenChainGraph compute node (treasury_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-42-arc-fit-diagnostic. Open at: https://ainumbers.co/chaingraph/art-46-arc-paymaster-model.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: ERC-4337 Paymaster economics model for Arc. Outputs: verdict, compliance_flags. Output schema: call describe_tool("model_arc_paymaster_economics").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
model_arc_stablefx_rfqArc StableFX RFQ Economics Model: OpenChainGraph compute node (treasury_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Arc StableFX RFQ Economics Model: OpenChainGraph compute node (treasury_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-42-arc-fit-diagnostic. Open at: https://ainumbers.co/chaingraph/art-44-arc-stablefx-model.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Quantify Herstatt risk elimination and FX spread savings from Arc StableFX 24/7 atomic PvP settlement vs non-CLS bilateral FX. Outputs: verdict, compliance_flags. Output schema: call describe_tool("model_arc_stablefx_rfq").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
model_buy_in_exposureBuy-In Exposure Modeler: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Buy-In Exposure Modeler: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-78-csdr-penalty-calculator. Output feeds: cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-83-buy-in-exposure-modeler.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Models CSDR Refit last-resort mandatory buy-in exposure: eligible trigger date per asset class, extension period (liquid equity ~7 cal days, gov bond ~12, SME ~22), buy-in cost mark-up (default 5%), and cash-compensation alternative. Outputs: buyin_triggered_count, modeled_fails, note, reference, status_note, total_buyin_exposure, total_cash_comp_exposure, total_fails. Output schema: call describe_tool("model_buy_in_exposure").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
model_cbam_certificate_costCBAM Certificate Cost & Free-Allocation Engine: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
CBAM Certificate Cost & Free-Allocation Engine: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-69-cbam-embedded-emissions-calculator. Output feeds: cry-05-agent-action-audit-trail-aggregator, art-76-climate-scenario-applicator. Open at: https://ainumbers.co/chaingraph/art-71-cbam-certificate-cost-engine.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Converts embedded emissions into a CBAM certificate liability: applies the CBAM factor (free-allocation phase-out 2.5% 2026 to 100% 2034), deducts origin carbon price already paid, and projects the quarterly holding and surrender schedule to the 30 Sep deadline. Outputs: cbam_factor, cbam_factor_applied, cbam_factor_year, certificate_liability_eur, certificates_required, eua_reference_price, free_allocation_phaseout_pct, gross_liability_tco2e, net_liability_eur, note, origin_price_credit, quarterly_holding_schedule, reference, surrender_deadline. Output schema: call describe_tool("model_cbam_certificate_cost").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
model_clearing_access_economicsClearing Access Model Selector: OpenChainGraph compute node (treasury_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Clearing Access Model Selector: OpenChainGraph compute node (treasury_mandate). Regulatory deadline: 2026-12-31 (flagship access-model decision (W-A).). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-48-treasury-clearing-fit-diagnostic. Output feeds: 504-settlement-risk-capital-optimizer, art-50-ficc-margin-netting-estimator. Open at: https://ainumbers.co/chaingraph/art-49-clearing-access-model-selector.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Selects and costs the FICC access model - Direct vs Sponsored (done-with) vs Sponsored/Agent (done-away) - across cost, execution-access, margin/netting efficiency, and ops. Outputs: annual_cost_by_model, cfo_memo, eligibility_gates, execution_access_score, im_estimate_by_model, model_scores, netting_efficiency_pct, note, recommended_model, segregation_recommendation. Output schema: call describe_tool("model_clearing_access_economics").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
model_perp_positionPerp Position Lifecycle: OpenChainGraph compute node (derivatives_margin_health).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Perp Position Lifecycle: OpenChainGraph compute node (derivatives_margin_health). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-213-perp-liquidation-calculator. Open at: https://ainumbers.co/chaingraph/art-214-perp-position-lifecycle.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Models a full perpetual futures position from open to close. Outputs: close_fee, close_fee_type, close_ts, disclaimer, entry_price, exit_price, funding_rate_per_interval, initial_margin, leverage, liq_price, maintenance_threshold, margin_return_pct, margin_returned, mmr_pct, n_intervals, notional, open_fee, open_fee_type, open_ts, position_size, price_delta, realized_pnl_gross, realized_pnl_net, side, total_fees, total_funding_impact, total_net_pnl, venue. Output schema: call describe_tool("model_perp_position").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
model_stablecoin_corridor_economicsStablecoin Corridor Economics Model: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Stablecoin Corridor Economics Model: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-249-compare-corridor-cost. Open at: https://ainumbers.co/chaingraph/art-250-model-stablecoin-corridor-economics.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Models the all-in cost of a USDC-based remittance corridor: on-ramp fee, chain gas fee, off-ramp/local-rail fee, FX spread, and pre-funding float savings vs correspondent banking. Outputs: break_even_usd, chain_fee_usd, correspondent_cost_pct, correspondent_cost_usd, disambiguation, float_savings_usd, fx_markup_share_of_traditional, fx_spread_usd, gross_cost_bps, gross_cost_pct, gross_stablecoin_cost_usd, meets_sdg_target, net_cost_bps, net_cost_pct, net_stablecoin_cost_usd, off_ramp_fee_usd, on_ramp_fee_usd, pii_note, regulatory_basis, savings_vs_traditional_pct, savings_vs_traditional_usd, sdg_target_pct, send_amount_usd, table_source, table_version. Output schema: call describe_tool("model_stablecoin_corridor_economics").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
model_tempo_gas_economicsTempo Fee-Sponsorship & Gas-AMM Economics: OpenChainGraph compute node (treasury_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Tempo Fee-Sponsorship & Gas-AMM Economics: OpenChainGraph compute node (treasury_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-35-tempo-payments-business-case. Open at: https://ainumbers.co/chaingraph/art-107-tempo-gas-economics.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Model Tempo enshrined-AMM gas cost paid in any major stablecoin, server-paid fee sponsorship, and net per-tx saving vs card/SWIFT/ACH baselines. Outputs: amm_slippage_bps, annual_saving, baseline_fee, blended_gas_cost, cfo_memo, effective_cost, per_tx_saving, server_paid_pct, sponsorship_breakeven_tx. Output schema: call describe_tool("model_tempo_gas_economics").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
model_x402_settlementx402 Settlement Cost & Finality Modeler: OpenChainGraph compute node (settlement_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
x402 Settlement Cost & Finality Modeler: OpenChainGraph compute node (settlement_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-01-ap2-mandate-chain-validator, art-12-acp-checkout-conformance-validator. Output feeds: art-30-agent-commerce-conformance-validator, art-61-x402-batch-settlement-reconciler, cry-04-merkle-batch-verifier, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-03-x402-settlement-modeler.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Rail-selection and finality recommendation across x402 (HTTP 402), Stripe USDC, card, ACH, and SWIFT. Outputs: eligible_rails, finality_sec, monthly_cost_usd, per_tx_fee_usd, recommended_rail. Output schema: call describe_tool("model_x402_settlement").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
optimize_settlement_capitalSettlement-Risk Capital Efficiency Optimizer: OpenChainGraph compute node (capital_assessment).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Settlement-Risk Capital Efficiency Optimizer: OpenChainGraph compute node (capital_assessment). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: 503-canton-tokenization-readiness-diagnostic. Open at: https://ainumbers.co/tools/504-settlement-risk-capital-optimizer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Quantify RWA and capital saving from Canton atomic DvP vs legacy settlement. Outputs: cet1_ratio, cost_of_capital, portfolio_bps, rows, total_annual_saving, total_capital_freed, total_notional_usd, total_rwa_delta_usd, verdict. Output schema: call describe_tool("optimize_settlement_capital").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
optimize_social_security_claim_ageSocial Security Claiming-Age Optimizer: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Social Security Claiming-Age Optimizer: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-283-pension-lump-sum-vs-annuity-decision-engine. Open at: https://ainumbers.co/chaingraph/art-282-social-security-claiming-optimizer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Models Social Security claiming-age tradeoffs from a claimant's own PIA/FRA statement figures: early-claim reduction and delayed-retirement-credit factors, earnings-test withholding below FRA, lifetime present value at 62/FRA/70/chosen age, and the 62-vs-70 undiscounted break-even age. Outputs: birthYear, breakEvenAge62vs70, claimAge, earningsTestAnnualLimit, earningsTestLimitSource, fullRetirementAge, lifetimePV, longevityAge, monthlyBenefitAtClaimAge, pia, recommendedClaimAge. Output schema: call describe_tool("optimize_social_security_claim_age").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
otlp_span_receiptOTLP span/trace to receipt bundle; does NOT build a ChainGraph (use build_chaingraph for that).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Description, as the server wrote it
OTLP span/trace to receipt bundle; does NOT build a ChainGraph (use build_chaingraph for that). Three modes, chosen by which input is supplied. (1) Pass `trace` alone to GENERATE a receipt bundle: a per-span receipt (span digest, parent-span receipt digest, semconv_snapshot, eddsa-jcs-2022 signature over an ephemeral did:key) for every span, plus a Merkle-rooted (RFC 6962) trace receipt. (2) Pass `bundle` ({trace, span_receipts, trace_receipt, issuer_did}) to VERIFY it: recomputes every digest, walks the parent-receipt chain, rebuilds the Merkle root, and reports which spans are attested, unattested, or tampered. (3) Pass `run_chain_result` (the structuredContent from run_chain) instead of `trace` to bridge a ChainGraph chain run into an OTLP trace first -- one execute_tool span per successfully-executed step, each carrying its own execution_hash as a span attribute (ocg.execution_hash) -- then generates the same receipt bundle over it. Byte-identical receipt engine to tools/566-otel-span-receipt-verifier.html -- a human's browser-generated bundle and this tool's output interoperate on the exact same wire shape.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: false
- openWorldHint: false
parse_camt053_reconciliationISO 20022 camt.053 Statement Reconciliation: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ISO 20022 camt.053 Statement Reconciliation: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-263-score-cash-forecast-accuracy. Open at: https://ainumbers.co/chaingraph/art-258-parse-camt053-reconciliation.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Classifies ISO 20022 camt.053 BkTxCd entries by Domain, Family, and SubFamily per the CGI-MP camt.053 Usage Guide v5.0 and the ISO 20022 ExternalBankTransactionCode1Code registry 2023-03. Outputs: balance_equation_passes, bucket_amounts, calculated_closing, closing_balance, credit_sum, day_count_convention, debit_sum, match_rate_pct, not_legal_advice, opening_balance, pii_note, reconciliation_status, regulatory_basis, structured_count, table_source, table_version, total_transactions, tx_counts_by_bucket, unstructured_count, variance. Output schema: call describe_tool("parse_camt053_reconciliation").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
plan_tls_pki_migrationTLS / X.509 PKI Migration Planner: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
TLS / X.509 PKI Migration Planner: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-85-pqc-timeline-fit-diagnostic, 499-crypto-asset-inventory-classifier. Output feeds: cry-04-merkle-batch-verifier, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-86-tls-pki-migration-planner.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Sequences TLS and X.509 PKI migration from RSA/ECDSA to post-quantum algorithms (ML-KEM/ML-DSA per NIST FIPS 203/204 Aug 2024). Outputs: algorithm_refs, estimated_total_weeks, interop_risks, inventory_ref, migration_plan, note, payload_impact_bytes, pki_summary, reference_version, rollback_points, strategy. Output schema: call describe_tool("plan_tls_pki_migration").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
precheck_reserve_attestationGENIUS Act Reserve Attestation Pre-Check: OpenChainGraph compute node (attestation_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
GENIUS Act Reserve Attestation Pre-Check: OpenChainGraph compute node (attestation_mandate). Regulatory deadline: 2027-01-01 (GENIUS Act effective ≤ January 2027; monthly reserve reports + annual PCAOB audit (>$50B issuers); FDIC NPRM April 2026). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-06-genius-act-reserve-attestation.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 1:1 reserve coverage, reserve-composition eligibility against GENIUS Act permitted-asset classes, monthly reserve-report figures against AICPA 2025 Criteria for CEO/CFO certification. Outputs: aicpa_2025_score_pct, applicable_deadline, asset_results, attestation_readiness_determination, conditional_assets_usd, coverage_ratio_pct, failing_dimensions, prohibited_assets_usd, regulatory_framework, reserve_shortfall_usd, total_liabilities_usd, total_reserves_usd, warnings. Output schema: call describe_tool("precheck_reserve_attestation").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
predict_settlement_failSettlement-Fail Predictor: OpenChainGraph compute node (model_governance).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Settlement-Fail Predictor: OpenChainGraph compute node (model_governance). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-77-t1-settlement-readiness-diagnostic, art-80-ssi-conformance-checker. Output feeds: art-84-settlement-efficiency-kpi, cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-79-settlement-fail-predictor.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Scores a trade's fail probability from anonymized configuration features (SSI match status, instrument liquidity tier, counterparty fail-history band, deadline proximity, partial-settlement availability) and ranks a batch for pre-settlement intervention. Outputs: batch_fail_rate_estimate, methodology, note, scored_trades, top_drivers, trade_count. Output schema: call describe_tool("predict_settlement_fail").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
prevalidation_readiness_scorerCross-Border Payment Prevalidation Readiness Scorer: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Cross-Border Payment Prevalidation Readiness Scorer: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-243-purpose-code-requirement-checker. Open at: https://ainumbers.co/chaingraph/art-247-prevalidation-readiness-scorer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Aggregate CBPR+ pre-validation readiness check for a single pacs.008 payment instruction. Outputs: cbpr_plus_deadline, chain_gate_note, check_details, checks_passed, checks_total, pii_note, readiness_pct, ready, regulatory_basis, remediation_actions, table_source, table_version. Output schema: call describe_tool("prevalidation_readiness_scorer").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
price_embedded_insuranceEmbedded Insurance Pricing Modeller: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Embedded Insurance Pricing Modeller: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-366-price-embedded-insurance.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Embedded-insurance unit economics for a platform attaching per-transaction coverage: per-transaction premium, monthly/annual gross written premium, net written premium after reinsurance cession, expected losses, commission and opex cost, underwriting profit, combined ratio, expense ratio, and bre... Outputs: annual_gwp, breakeven_loss_ratio_pct, combined_ratio_pct, commission_cost, expected_losses, expense_ratio_pct, monthly_gwp, net_written_premium, note, opex_cost, per_tx_premium, underwriting_profit. Output schema: call describe_tool("price_embedded_insurance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
prove_metadata_sanitizationMetadata Sanitization Prover: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Metadata Sanitization Prover: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-191-conversion-receipt-builder. Open at: https://ainumbers.co/chaingraph/art-193-metadata-sanitization-prover.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Produces a proof-of-sanitization record binding the original digest to the findings removed, redacted, or retained and to the sanitized digest, with a deterministic residual-risk analysis by file type (JPEG COM segments, PNG textual chunks, PDF and DOCX XMP and embedded-object metadata). Outputs: residual_risks, sanitization_record, verdict. Output schema: call describe_tool("prove_metadata_sanitization").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
recompute_fund_navRecompute Fund NAV: OpenChainGraph compute node (attestation_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Recompute Fund NAV: OpenChainGraph compute node (attestation_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-373-recompute-fund-nav.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Recomputes a fund's net-asset-value-per-share from SUPPLIED holdings (quantity x supplied price, multi-currency with supplied FX), accruals (income and expense with a declared day-count convention), liabilities, and shares outstanding, using fixed-point BigInt money math throughout (no float accu... Outputs: base_currency, components, fence, fund_id, nav_per_share, not_proven, regulatory_framework, rounding, structural_error, valuation_date. Output schema: call describe_tool("recompute_fund_nav").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
recon_matchParses a camt.053.001 statement (same schema-subset checks as camt053_parse) and a counterpart expectation set (CSV, header EndToEndId,Amount,Currency,Date -- same strict RFC 4180 …No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Parses a camt.053.001 statement (same schema-subset checks as camt053_parse) and a counterpart expectation set (CSV, header EndToEndId,Amount,Currency,Date -- same strict RFC 4180 parser as the WORKBOOK-1 CSV tools), then runs the SAME deterministic match engine as the tools/565 browser workbench: EndToEndId exact match first (statement order), then amount+date tolerance + currency match (first unmatched expectation in CSV order). Returns matches, exceptions (unmatched entries/expectations), and a reconciliation receipt (statement digest, expectation-set digest, match-rule declaration, counts, exception digests, execution_hash) -- byte-identical to the browser tool for the same inputs. Prepare/hash/receipt only; per-exception disposition receipts are the browser workbench's interactive follow-on.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
reconcile_commission_statementCommission Statement Reconciler: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Commission Statement Reconciler: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-264-validate-commission-hierarchy. Output feeds: art-265-amortize-asc606-commissions. Open at: https://ainumbers.co/chaingraph/art-266-reconcile-commission-statement.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Line-item reconciliation of expected vs. Outputs: discrepancy_amount, discrepancy_classification, discrepancy_pct, has_discrepancy, line_count, line_results, not_legal_advice, pii_note, regulatory_basis, table_source, table_version, tolerance_pct, total_expected, total_stated. Output schema: call describe_tool("reconcile_commission_statement").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
reconcile_emir_pairingEMIR Counterparty Pairing Reconciler: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
EMIR Counterparty Pairing Reconciler: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-157-emir-lifecycle-event-validator. Open at: https://ainumbers.co/chaingraph/art-156-emir-counterparty-pairing-reconciler.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
reconcile_erc8056_multiplierERC-8056 Multiplier Reconciler: OpenChainGraph compute node (collateral_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ERC-8056 Multiplier Reconciler: OpenChainGraph compute node (collateral_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-317-rhc-multiplier-reconciler.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Reconciles Robinhood Chain stock-token corporate actions against the ERC-8056 scaled UI amount surface. Outputs: computed_ratio, declared_action_type, discrepancies, event_count, expected_ratio, ratio_match, raw_balance_invariant, verdict. Output schema: call describe_tool("reconcile_erc8056_multiplier").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
reconcile_mpp_subscriptionTempo Subscription & Streaming Settlement Reconciler: OpenChainGraph compute node (settlement_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Tempo Subscription & Streaming Settlement Reconciler: OpenChainGraph compute node (settlement_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-36-tempo-mpp-agent-mandate. Output feeds: cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-106-tempo-subscription-reconciler.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Reconcile executed MPP recurring/streamed draws against the authorized mandate envelope, prove draw-set integrity via Merkle root, and detect per-cycle cap breaches, cumulative-cap violations, and mandate-expiry/revocation breaches. Outputs: breaches, cumulative_ok, cycles_ok, draw_count, draw_merkle_root, residual_envelope, total_drawn, verdict. Output schema: call describe_tool("reconcile_mpp_subscription").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
reconcile_sii_ifrs17SII-IFRS 17 Reconciliation Bridger: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
SII-IFRS 17 Reconciliation Bridger: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-180-solvency2-scr-ratio-calculator. Output feeds: art-182-insurance-reporting-readiness-diagnostic. Open at: https://ainumbers.co/chaingraph/art-181-sii-ifrs17-reconciliation-bridger.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Bridge Solvency II technical provisions (best estimate + risk margin) to IFRS 17 insurance contract liabilities (fulfilment cash flows + risk adjustment + CSM). Outputs: bel_fcf_delta, bridge_delta, bridge_within_tolerance, ifrs17_csm, ifrs17_fcf, ifrs17_insurance_contract_liabilities, ifrs17_ra, ra_vs_risk_margin_ratio_pct, relative_bridge_delta_pct, rm_ra_delta, sii_best_estimate, sii_risk_margin, sii_technical_provisions. Output schema: call describe_tool("reconcile_sii_ifrs17").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
reconcile_x402_batch_settlementx402 V2 Batch-Settlement Reconciler: OpenChainGraph compute node (settlement_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
x402 V2 Batch-Settlement Reconciler: OpenChainGraph compute node (settlement_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-03-x402-settlement-modeler, art-60-agent-economy-runtime-fit-diagnostic. Output feeds: art-62-ap2-payment-receipt-verifier, cry-04-merkle-batch-verifier, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-61-x402-batch-settlement-reconciler.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Reconciles an x402 V2 batch settlement (off-chain payment vouchers vs onchain batch total), verifying recon verdict, per-voucher amounts, settlement-risk window (unredeemed voucher exposure), and computing an educational Merkle root over the voucher set. Outputs: batch_delta_minor_units, batch_id, escrow_address, merkle_root_preimage, note, recon_verdict, redeemed_count, redeemed_total, settlement_asset, settlement_risk_window, status_asof, unredeemed_value, voucher_count, voucher_findings, within_tolerance. Output schema: call describe_tool("reconcile_x402_batch_settlement").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
redline_verifyRecomputes a redline review from scratch: the diff (from the pasted original+revised text), the diff_receipt's execution_hash, every hunk_receipt's digest/execution_hash/chain-link…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Recomputes a redline review from scratch: the diff (from the pasted original+revised text), the diff_receipt's execution_hash, every hunk_receipt's digest/execution_hash/chain-link (in the order supplied -- entries may be minted by an agent, a human, or both interleaved), the disposition_receipt's execution_hash, and the accepted-text digest. Returns valid:true/false, a per-check breakdown, and broken_hunk (the zero-based index of the first divergence, or null) -- never trusts a bundle claim without checking it against a fresh recomputation.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
replay_supervisory_scenarioSupervisory Scenario Replay (DFAST-lite): OpenChainGraph compute node (capital_assessment).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Supervisory Scenario Replay (DFAST-lite): OpenChainGraph compute node (capital_assessment). Regulatory deadline: 2027-02-01 (Annual re-pin — Fed publishes new supervisory scenarios each February). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: sim-01-lcr-nsfr-liquidity-stress-test, sim-03-basel-rwa-scenario-modeler. Open at: https://ainumbers.co/chaingraph/art-370-supervisory-scenario-replay.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Replays the Fed's published 2026 28-variable supervisory scenario paths (baseline and severely adverse, Q1:2026-Q1:2029) against user-supplied loss and PPNR coefficient functions, producing a quarterly P&L and capital walk. Outputs: ending_capital_mn, not_a_submission, quarters, scenario, scenario_release_date, scenario_set_digest, scenario_source_url, starting_capital_mn, trough_capital_mn, trough_capital_ratio_pct, trough_quarter. Output schema: call describe_tool("replay_supervisory_scenario").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
rollforward_y14_capital_worksheetFR Y-14 Capital Worksheet Roll-Forward & Cross-Check: OpenChainGraph compute node (regulatory_reporting).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
FR Y-14 Capital Worksheet Roll-Forward & Cross-Check: OpenChainGraph compute node (regulatory_reporting). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-439-y14-capital-worksheet-rollforward.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Rolls forward a caller-declared FR Y-14A/Q capital worksheet (CET1, additional Tier 1, Tier 2) from beginning balance through period additions and deductions to an ending balance, applies a caller-declared published-scenario adjustment (e.g. Outputs: boundary_note, constants_version, cross_check, ending_tier1_capital_usd, ending_total_capital_usd, entity_id, published_scenario, report_form, reporting_period, rollforward, worksheet. Output schema: call describe_tool("rollforward_y14_capital_worksheet").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
route_einvoice_jurisdiction_mandateE-Invoice Jurisdiction Mandate Router: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
E-Invoice Jurisdiction Mandate Router: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-294-einvoice-vat-calc-verifier. Output feeds: art-296-einvoice-transmission-receipt-builder. Open at: https://ainumbers.co/chaingraph/art-295-einvoice-jurisdiction-mandate-router.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Deterministic lookup over a version-pinned mandate table: given supplier/buyer country, transaction type, and transaction date, routes to the applicable e-invoicing regime, format, phase status, and transmission channel. Outputs: applicable_format, mandatory_from, phase_status, regime_country, table_version, transmission_channel. Output schema: call describe_tool("route_einvoice_jurisdiction_mandate").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
route_partner_stablecoin_jurisdictionArc Multi-Currency Corridor Jurisdiction Router: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Arc Multi-Currency Corridor Jurisdiction Router: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: 511-multi-currency-pvp-validator. Open at: https://ainumbers.co/chaingraph/art-111-arc-corridor-jurisdiction-router.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Route each leg of a multi-currency Arc corridor to its per-currency home regime (EURC→MiCA EMT, JPYC→JP FSA, BRLA→Brazil CMN/BCB, MXNB→MX CNBV, etc.), flag missing per-regime disclosures, and emit PvP handoff (511) + Travel-Rule batch handoff (art-104). Outputs: disclosure_gaps, leg_regimes, pvp_handoff, travel_rule_handoff, verdict. Output schema: call describe_tool("route_partner_stablecoin_jurisdiction").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
route_vida_oss_registrationViDA OSS Registration Router: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
ViDA OSS Registration Router: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-162-vida-platform-deemed-supplier-classifier. Output feeds: art-164-vida-compliance-readiness-diagnostic. Open at: https://ainumbers.co/chaingraph/art-163-vida-oss-registration-router.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_401k_adp_acp_test401(k) ADP/ACP Nondiscrimination Tester: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "run_401k_adp_acp_test" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_401k_adp_acp_test" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_401k_adp_acp_test" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
401(k) ADP/ACP Nondiscrimination Tester: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-302-401k-adp-acp-test.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Runs the IRC §401(k)(3) Actual Deferral Percentage test and the §401(m)(2) Actual Contribution Percentage test from supplied HCE vs NHCE deferral/match percentages (current-year or prior-year method), applying the fixed statutory permitted-disparity limits (1.25x, or 2 percentage points and 2x, w... Outputs: acp, adp, all_tests_pass, error, method. Output schema: call describe_tool("run_401k_adp_acp_test").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_agent_economy_fitAgent Economy Runtime Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_agent_economy_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_agent_economy_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_agent_economy_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Agent Economy Runtime Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-61-x402-batch-settlement-reconciler, art-62-ap2-payment-receipt-verifier, art-63-agent-service-metering-modeler, art-02-agent-spend-policy-simulator, mms-03-app-fraud-graph, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-60-agent-economy-runtime-fit-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 12-question A-F readiness diagnostic for the agent-economy runtime / post-trade layer (x402 V2 batch settlement, AP2 PaymentReceipt, Human-Not-Present autonomy, reconciliation, metering, runtime fraud). Outputs: dim_scores, hnp_risk_flag, note, overall_grade, overall_score, primary_recommendation, remediation_checklist, secondary_recommendations. Output schema: call describe_tool("run_agent_economy_fit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_agentic_readiness_diagnosticAgentic Payments Readiness Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_agentic_readiness_diagnostic" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_agentic_readiness_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_agentic_readiness_diagnostic" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Agentic Payments Readiness Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-15-agentic-mandate-sandbox, art-16-google-ap2-mandate-builder, art-17-ap2-mcp-policy-validator, art-18-mcp-developer-readiness-scorecard. Open at: https://ainumbers.co/chaingraph/art-27-agentic-readiness-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_ai_act_highrisk_fitEU AI Act High-Risk Fit & Classification Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_ai_act_highrisk_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_ai_act_highrisk_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_ai_act_highrisk_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
EU AI Act High-Risk Fit & Classification Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-65-ai-conformity-pack-builder, art-66-fria-postmarket-monitoring-builder, art-67-agentic-ai-risk-classifier, art-05-eu-ai-act-credit-scoring-conformity, 452-fair-lending-ai-bias-assessment, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-64-ai-act-highrisk-fit-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 12-question A-F diagnostic that screens in-force obligations first (Art 5 prohibited practices, Art 4 AI literacy, GPAI) then classifies financial AI system high-risk status (Annex III: credit scoring, insurance pricing, financial-standing) and grades Article 9-15 readiness plus the provider/depl... Outputs: ai_literacy_grade, annex_iii_basis, applicable_date, dim_scores, do_now_checklist, gpai_applicability, high_risk_verdict, note, overall_grade, overall_score, prepare_ahead_checklist, primary_recommendation, prohibited_practice_verdict, role, secondary_recommendations. Output schema: call describe_tool("run_ai_act_highrisk_fit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_ai_governance_fitAI Governance Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "run_ai_governance_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_ai_governance_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_ai_governance_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
AI Governance Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-175-gpai-code-of-practice-conformance. Open at: https://ainumbers.co/chaingraph/art-176-ai-governance-readiness-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: A-F AI governance readiness diagnostic across six dimensions spanning ISO/IEC 42001, NIST AI RMF, and EU AI Act convergence: AIMS documentation, impact assessment, NIST RMF mapping, GPAI obligation check, monitoring plan, and review cycle. Outputs: dimensions_met, dimensions_total, frameworks_addressed, fully_ready, gaps, readiness_grade, readiness_score. Output schema: call describe_tool("run_ai_governance_fit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_arc_fit_diagnosticArc Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_arc_fit_diagnostic" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_arc_fit_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_arc_fit_diagnostic" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Arc Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-43-arc-cpn-model, art-44-arc-stablefx-model, art-45-arc-xreserve-linter, art-46-arc-paymaster-model, art-47-arc-cctp-transfer. Open at: https://ainumbers.co/chaingraph/art-42-arc-fit-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_call_report_edit_checksCall Report Published Edit-Check Gate: OpenChainGraph compute node (regulatory_reporting).1 finding
Why it is in this group
- The name "run_call_report_edit_checks" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_call_report_edit_checks" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_call_report_edit_checks" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Call Report Published Edit-Check Gate: OpenChainGraph compute node (regulatory_reporting). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-432-call-report-rc-balance-sheet, art-433-call-report-rcr-capital. Open at: https://ainumbers.co/chaingraph/art-434-call-report-edit-check-gate.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Runs a curated battery of FFIEC-style published Call Report edit checks -- balance-sheet identity, capital-stack ordering (CET1 <= Tier 1 <= Total capital), ratio-vs-component consistency, cross-schedule entity/period match -- against art-432 (Schedule RC) and art-433 (Schedule RC-R) output paylo... Outputs: all_fatal_passed, check_count, checks, coverage_note, entity_id, fatal_failure_count, gate_status, report_form, reporting_period, rounding_tolerance_usd, warning_failure_count. Output schema: call describe_tool("run_call_report_edit_checks").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_carbon_compliance_fitCarbon & Climate Compliance Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_carbon_compliance_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_carbon_compliance_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_carbon_compliance_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Carbon & Climate Compliance Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-69-cbam-embedded-emissions-calculator, art-72-cbam-precursor-emissions-aggregator, art-73-taxonomy-alignment-scorer, art-75-eugb-factsheet-validator, art-76-climate-scenario-applicator, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-68-carbon-compliance-fit-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 12-question A-F diagnostic that classifies which carbon/climate obligations bind a firm (CBAM authorised-declarant duty, EU Taxonomy alignment, EU Green Bond conformance, climate stress) and routes to the right carbon-compliance chain. Outputs: cbam_declarant_required, cbam_declarant_verdict, cbam_dual_status, climate_stress_applicable, dim_scores, do_now_checklist, eugb_readiness, note, overall_grade, overall_score, prepare_ahead_checklist, primary_recommendation, secondary_recommendations, taxonomy_scope. Output schema: call describe_tool("run_carbon_compliance_fit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_chainExecutes every step of a named chain (list names with find_chain / build_workflow_links) and returns ONE composite artifact whose execution_hash anchors all step outputs.1 finding
Why it is in this group
- The name "run_chain" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_chain" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_chain" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Executes every step of a named chain (list names with find_chain / build_workflow_links) and returns ONE composite artifact whose execution_hash anchors all step outputs. compute:"server"/"auto" (default) runs each kernel-backed step server-side, threading step N's execution_hash into step N+1's parent_hashes; compute:"browser" returns a zero-egress delegation bundle (composer URL + ordered deep-links) to run client-side instead — no data leaves the agent. Supply inputs as a map of step tool_id -> policy_parameters (field names per node manifest / build_chaingraph); a step whose kernel needs inputs you omit is reported per-step (status "input_required"), never failed silently. Steps that are browser-only (gpu:true or no registered kernel) are listed for browser delegation. Deterministic, zero PII, zero payload logging. Verify the result with verify_execution_hash. Runs with anything to explain carry decision_trail: per-step reason codes (gate rule id, escalation rule id, input_required cause), hash-excluded adjacent metadata recomputable from the hash-bound decisions[]. Each server-mode run also returns an OpenTelemetry GenAI span document as a resource link (one execute_tool span per executed step under an invoke_agent parent). Response includes a ledger_url fragment link for human verification at ledger.ainumbers.co. Server-mode responses also echo a stable `dedupe` object — `dedupe.input_hash` (JCS-SHA-256 over the effective run inputs: chain, per-step inputs after the caller/fixture fallback, and mandate_hash when a mandate governs) beside `dedupe.composite_execution_hash` — so a client can recognize and skip an exact re-run (idempotentHint) without recomputing anything. See docs/IDEMPOTENCY.md.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_chain_batchBatch over named chains in ONE round-trip with a PER-ROW terminal status — one row's failure never fails another, and every row's result rides the single response.1 finding
Why it is in this group
- The name "run_chain_batch" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_chain_batch" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_chain_batch" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Batch over named chains in ONE round-trip with a PER-ROW terminal status — one row's failure never fails another, and every row's result rides the single response. mode:"run" (default) executes each row through the same engine as run_chain (composite artifact + execution_hash per row, deterministic); mode:"estimate" validates every row (chain exists, step counts, compute feasibility, inputs coverage, OCG §21.4 gate rule previews) WITHOUT executing anything. Capped at 8 rows and 11 total server-kernel steps per batch — caps MEASURED live on the deployed free-plan endpoint, where N kernel builds in one invocation share one CPU budget (an 11-step chain: green at 416 ms). Beyond the cap, use sequential run_chain calls — each call gets its own CPU budget (the blessed fan-out pattern on this tier; docs/client-fan-out-patterns.md). Retry = re-run the WHOLE batch: safe, because determinism reproduces byte-identical execution hashes. Mandates and input_required escalation are not accepted on batch rows; escalated rows carry the non-blocking resolve_handle record. Zero PII, zero payload logging.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_digital_trade_fitDigital Trade Corridor Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_digital_trade_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_digital_trade_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_digital_trade_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Digital Trade Corridor Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-53-mletr-ebl-conformance-validator, art-54-digital-trade-rules-checker, art-55-trade-document-provenance-verifier, 509-canton-party-allowlist-validator, art-10-amla-transaction-typology-risk-scorer, ml-02-credit-default-risk-scorer, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-52-digital-trade-fit-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 12-question A–F readiness diagnostic for digital trade / electronic trade documents (MLETR). Outputs: corridor_enforceability_flag, dim_scores, note, overall_grade, overall_score, primary_recommendation, remediation_checklist, secondary_recommendations. Output schema: call describe_tool("run_digital_trade_fit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_dora_readiness_diagnosticDORA Readiness Diagnostic: OpenChainGraph compute node (infrastructure_mandate).1 finding
Why it is in this group
- The name "run_dora_readiness_diagnostic" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_dora_readiness_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_dora_readiness_diagnostic" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
DORA Readiness Diagnostic: OpenChainGraph compute node (infrastructure_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-09-dora-incident-classifier, pnr-01-dora-ict-cascade-simulator, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-29-dora-readiness-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 12-question scored diagnostic across four DORA pillars (ICT risk management, incident classification & reporting, resilience testing, third-party risk). Outputs: all_answered, applicable_deadline_note, domain_scores, gaps, gaps_count, grade, grade_title, immediate_action_required, incident_route_recommended, regulatory_framework, score_pct, supervisory_exposure. Output schema: call describe_tool("run_dora_readiness_diagnostic").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_emir_reporting_fitEMIR Reporting Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "run_emir_reporting_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_emir_reporting_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_emir_reporting_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
EMIR Reporting Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-157-emir-lifecycle-event-validator. Open at: https://ainumbers.co/chaingraph/art-158-emir-reporting-readiness-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_eudr_readiness_fitEUDR Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "run_eudr_readiness_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_eudr_readiness_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_eudr_readiness_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
EUDR Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-169-eudr-supply-chain-traceability-linker. Open at: https://ainumbers.co/chaingraph/art-170-eudr-readiness-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: A-F EUDR readiness diagnostic across six dimensions: scope mapping, geolocation data quality, DDS submission readiness (TRACES NT), country risk assessment, risk mitigation documentation, and 5-year retention system. Outputs: dimensions_met, dimensions_total, enforcement_deadlines, fully_ready, gaps, readiness_grade, readiness_score. Output schema: call describe_tool("run_eudr_readiness_fit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_illustration_selfsupport_testLife Illustration Self-Support Test (NAIC Model 582): OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "run_illustration_selfsupport_test" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_illustration_selfsupport_test" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_illustration_selfsupport_test" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Life Illustration Self-Support Test (NAIC Model 582): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-254-compute-rbc-action-level. Open at: https://ainumbers.co/chaingraph/art-253-run-illustration-selfsupport-test.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Runs the NAIC Model Regulation 582 §8C self-support test (year 15 and year 20 account value positive) and §8D lapse-support prohibition check for life insurance illustrations. Outputs: account_value_yr15, account_value_yr20, cumulative_net, face_amount, illustration_valid, issues, lapse_adjusted_av_yr20, lapse_support_flag, not_legal_advice, pii_note, policy_years_provided, regulatory_basis, self_support_pass, self_support_yr15_pass, self_support_yr20_pass, table_source, table_version, total_coi, total_expenses, total_interest, total_premiums. Output schema: call describe_tool("run_illustration_selfsupport_test").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_insurance_reporting_fitInsurance Reporting Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "run_insurance_reporting_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_insurance_reporting_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_insurance_reporting_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Insurance Reporting Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-181-sii-ifrs17-reconciliation-bridger. Open at: https://ainumbers.co/chaingraph/art-182-insurance-reporting-readiness-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: A-F insurance reporting readiness diagnostic across six dimensions: IFRS 17 measurement model election, CSM system implementation, risk-adjustment disclosure, Solvency II Pillar-3 QRT reporting, SII-IFRS 17 reconciliation, and ICS assessment for IAIGs. Outputs: dimensions_met, dimensions_total, fully_ready, gaps, readiness_grade, readiness_score. Output schema: call describe_tool("run_insurance_reporting_fit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_irrbb_disclosure_fitIRRBB Disclosure Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "run_irrbb_disclosure_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_irrbb_disclosure_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_irrbb_disclosure_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
IRRBB Disclosure Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-187-irrbb-csrbb-scope-checker. Open at: https://ainumbers.co/chaingraph/art-188-irrbb-disclosure-readiness-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: A-F IRRBB disclosure readiness diagnostic across five dimensions: EVE shock calculation performed, SOT (EVE + NII) evaluated, standardised approach mapped, CSRBB scope assessed, and Pillar 3 IRRBB1 disclosure template ready. Outputs: dimensions_met, dimensions_total, fully_ready, gaps, readiness_grade, readiness_score. Output schema: call describe_tool("run_irrbb_disclosure_fit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_kernel_vmRun a ChainGraph decision kernel's compute(policy_parameters) inside a sandboxed, deterministic, in-browser QuickJS-ng WebAssembly VM (ocg-deterministic-compute@2) and return its o…1 finding
Why it is in this group
- The name "run_kernel_vm" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_kernel_vm" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_kernel_vm" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Run a ChainGraph decision kernel's compute(policy_parameters) inside a sandboxed, deterministic, in-browser QuickJS-ng WebAssembly VM (ocg-deterministic-compute@2) and return its output_payload. Demo kernel set only -- for the full catalog, use the worker's compute kernels directly. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("run_kernel_vm").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_mcp_deployability_diagnosticMCP Server Deployability Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_mcp_deployability_diagnostic" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_mcp_deployability_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_mcp_deployability_diagnostic" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
MCP Server Deployability Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-18-mcp-developer-readiness-scorecard. Open at: https://ainumbers.co/chaingraph/art-28-mcp-server-deployability-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Output schema: call describe_tool("run_mcp_deployability_diagnostic").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_mica_casp_fitMiCA CASP Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_mica_casp_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_mica_casp_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_mica_casp_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
MiCA CASP Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-99-mica-transitional-deadline-router, art-100-mica-casp-authorization-readiness, art-102-crypto-asset-whitepaper-linter, art-103-mar-crypto-surveillance-readiness, art-104-tfr-travel-rule-batch-validator, art-105-mica-token-service-scoper, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-98-mica-casp-fit-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 12-question A-F diagnostic scoping a crypto-asset service provider's MiCA Title-V lifecycle readiness (authorization, Art 67 own-funds, whitepaper, MAR-crypto, travel rule) and routing to the right MiCA chain. Outputs: dim_scores, gaps, note, primary_recommendation, readiness_grade, reference_version, secondary_recommendations, services_count. Output schema: call describe_tool("run_mica_casp_fit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_pqc_timeline_fitPQC Timeline & Migration Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_pqc_timeline_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_pqc_timeline_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_pqc_timeline_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
PQC Timeline & Migration Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-86-tls-pki-migration-planner, art-87-iso20022-pqc-readiness-checker, art-88-fido-pqc-conformance-checker, art-89-blockchain-quantum-risk-classifier, 499-crypto-asset-inventory-classifier, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-85-pqc-timeline-fit-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 12-dimension A-F diagnostic mapping an organisation's cryptographic estate and sector to the CNSA 2.0 / EU-2030 / G7 / DORA post-quantum milestones, flagging the end-2026 EU crypto-inventory deadline. Outputs: dim_scores, do_now, inventory_deadline_flag, milestone_fit, milestones, note, prepare_ahead, primary_recommendation, protocol_routes, readiness_grade, reference_version, total_score. Output schema: call describe_tool("run_pqc_timeline_fit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_rate_shock_ladderRate Shock Ladder Replay: OpenChainGraph compute node (analytics_mandate).1 finding
Why it is in this group
- The name "run_rate_shock_ladder" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_rate_shock_ladder" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_rate_shock_ladder" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Rate Shock Ladder Replay: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-369-run-rate-shock-ladder.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: US OCC/FDIC interest-rate-risk parallel shock ladder: sweeps four prescribed parallel magnitudes (+/-100/200/300/400bp) over a bucketed repricing-gap schedule, returning both a duration-weighted delta-EVE leg and a 12-month cumulative-gap delta-NII leg per shock. Outputs: buckets_used, convention, ladder, magnitudes_bps, nii_12m_gap, preset_shocks, total_net_gap, worst_delta_eve, worst_scenario. Output schema: call describe_tool("run_rate_shock_ladder").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_robinhood_chain_fit_diagnosticRobinhood Chain Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_robinhood_chain_fit_diagnostic" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_robinhood_chain_fit_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_robinhood_chain_fit_diagnostic" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Robinhood Chain Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-317-rhc-multiplier-reconciler, art-318-rhc-regime-mapper, art-319-rhc-valuation-linter, art-320-rhc-collateral-haircut, art-321-rhc-bold-finality-classifier, art-322-rhc-ap-redemption-stress. Open at: https://ainumbers.co/chaingraph/art-323-rhc-fit-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_sanctions_screening_fitSanctions & Export-Control Screening Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_sanctions_screening_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_sanctions_screening_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_sanctions_screening_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Sanctions & Export-Control Screening Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-91-ownership-50pct-aggregator, art-92-screening-list-coverage-checker, art-93-fuzzy-match-calibration-scorer, art-94-eccn-dual-use-classifier, art-95-circumvention-diligence-assessor, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-90-sanctions-screening-fit-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 12-param A-F diagnostic scoping a firm's sanctions/export-control screening program (50%-rule ownership, list coverage, fuzzy-match calibration, ECCN classification, circumvention controls) and routing to the right sanctions/export-control chain. Outputs: always_route, dim_scores, gaps, key_dates, note, primary_recommendation, program_grade, raw_score, reference_version, secondary_routes. Output schema: call describe_tool("run_sanctions_screening_fit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_section125_ndt§125 Cafeteria Plan Nondiscrimination Tester: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "run_section125_ndt" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_section125_ndt" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_section125_ndt" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
§125 Cafeteria Plan Nondiscrimination Tester: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-301-section125-ndt.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Runs the §125 cafeteria-plan nondiscrimination tests from supplied aggregate participant counts: an eligibility-ratio test and a contributions-and-benefits ratio test (§410(b)-style ratio-percentage analogy -- IRS has never finalized regs under §125(g)(3), DRAFT pending final rules), and the stat... Outputs: all_tests_pass, benefits, concentration, eligibility, error. Output schema: call describe_tool("run_section125_ndt").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_t1_readiness_diagnosticT+1 Settlement Readiness Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_t1_readiness_diagnostic" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_t1_readiness_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_t1_readiness_diagnostic" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
T+1 Settlement Readiness Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-78-csdr-penalty-calculator, art-79-settlement-fail-predictor, art-80-ssi-conformance-checker, art-81-allocation-affirmation-conformance, art-82-securities-settlement-message-linter, art-83-buy-in-exposure-modeler, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-77-t1-settlement-readiness-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 12-question A-F diagnostic scoring a firm's readiness for the coordinated EU/UK/CH T+1 move (11 Oct 2027) against the Industry Roadmap phases, grading trade-date allocation/confirmation (the Dec-2026 23:00 CET machine-readable mandate), SSI automation, funding compression and CSDR-penalty exposur... Outputs: binding_deadline, dim_scores, dual_date_note, firm_type, gap_checklist, jurisdictions, note, overall_score, primary_recommendation, readiness_grade, secondary_recommendations. Output schema: call describe_tool("run_t1_readiness_diagnostic").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_tempo_fit_diagnosticTempo Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_tempo_fit_diagnostic" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_tempo_fit_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_tempo_fit_diagnostic" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Tempo Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-35-tempo-payments-business-case, art-36-tempo-mpp-agent-mandate, art-37-tempo-stablecoin-issuance, art-40-tempo-agentic-checkout. Open at: https://ainumbers.co/chaingraph/art-34-tempo-fit-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_tokenized_settlement_fitWholesale Tokenized Settlement Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_tokenized_settlement_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_tokenized_settlement_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_tokenized_settlement_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Wholesale Tokenized Settlement Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-57-deposit-token-compliance-validator, art-58-cross-network-settlement-validator, art-59-settlement-asset-finality-classifier, 505-tokenized-collateral-eligibility-checker, 509-canton-party-allowlist-validator, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-56-tokenized-settlement-fit-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 12-question A-F readiness diagnostic for wholesale tokenized settlement (tokenized deposits, central bank money, regulated stablecoins as settlement assets). Outputs: dim_scores, finality_flag, note, overall_grade, overall_score, primary_recommendation, remediation_checklist, secondary_recommendations. Output schema: call describe_tool("run_tokenized_settlement_fit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_treasury_clearing_fitTreasury Clearing Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_treasury_clearing_fit" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_treasury_clearing_fit" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_treasury_clearing_fit" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
Treasury Clearing Fit Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Regulatory deadline: 2026-12-31 (SEC UST clearing: cash Dec 31 2026, repo Jun 30 2027. D0 root of all treasury-clearing chains.). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-49-clearing-access-model-selector, art-50-ficc-margin-netting-estimator. Open at: https://ainumbers.co/chaingraph/art-48-treasury-clearing-fit-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 12-question A-F readiness diagnostic for the SEC US Treasury clearing mandate (cash Dec 31 2026 / repo Jun 30 2027). Outputs: compliance_deadline, dim_scores, note, overall_grade, overall_score, primary_recommendation, remediation_checklist, secondary_recommendations. Output schema: call describe_tool("run_treasury_clearing_fit").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_umr_aana_readinessUMR / AANA Readiness Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate).1 finding
Why it is in this group
- The name "run_umr_aana_readiness" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_umr_aana_readiness" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_umr_aana_readiness" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
UMR / AANA Readiness Diagnostic: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-407-umr-aana-readiness-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Determines whether a group's declared AANA (average aggregate notional amount) puts it in scope for the uncleared margin rules (UMR), per AT-CLEARING-WAVE-SPEC.md CW-2, and flags which counterparties are over the regulatory initial-margin threshold with an open documentation or custody gap. Outputs: aana_group_eur, aana_in_scope_threshold_eur, constants_version, counterparties, counterparties_over_im_threshold, counterparty_count, disambiguation, im_threshold_eur, in_scope_aana, note, overall_grade, remediation_checklist, thresholds_vintage. Output schema: call describe_tool("run_umr_aana_readiness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
run_vida_readiness_diagnosticViDA Compliance Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "run_vida_readiness_diagnostic" says "run": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "run_vida_readiness_diagnostic" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "run_vida_readiness_diagnostic" says "run": a word for a change.
  What this check looks for
Description, as the server wrote it
ViDA Compliance Readiness Diagnostic: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-163-vida-oss-registration-router. Open at: https://ainumbers.co/chaingraph/art-164-vida-compliance-readiness-diagnostic.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
scan_tool_poisoningScan an MCP tool description/manifest for tool-poisoning and prompt-injection smells; returns a risk score and flagged patterns.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Scan an MCP tool description/manifest for tool-poisoning and prompt-injection smells; returns a risk score and flagged patterns. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("scan_tool_poisoning").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
scope_mica_token_and_serviceMiCA Token & Service Scoper: OpenChainGraph compute node (agent_guardrail_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
MiCA Token & Service Scoper: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-98-mica-casp-fit-diagnostic. Output feeds: cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-105-mica-token-service-scoper.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Disambiguation router classifying a case as ART/EMT-issuer (delegated to existing stablecoin-compliance chains) vs CASP-service (MiCA chains). Outputs: classification, delegated_to_existing, existing_chains_delegated, mica_note, rationale, reference_version, route_target, wave20_chains_applicable. Output schema: call describe_tool("scope_mica_token_and_service").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_agent_insurability_evidenceAgent Insurability Evidence Scorer: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Agent Insurability Evidence Scorer: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-307-claim-dispute-bundle-builder. Open at: https://ainumbers.co/chaingraph/art-306-agent-insurability-evidence-scorer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Scores an agent execution evidence bundle for underwriter-facing evidence completeness across four dimensions (determinism, replayability, oversight density, dispute history) using a version-pinned rubric selected by underwriter_profile: aiuc, aisure (Munich Re aiSure evidence-doc list), armilla... Outputs: composite, dims, incident_history_provided, insufficient_evidence, reputation_self_asserted, rubric_version, underwriter_profile. Output schema: call describe_tool("score_agent_insurability_evidence").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_aml_typologiesAMLA Transaction-Typology Risk Scorer: OpenChainGraph compute node (risk_control).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
AMLA Transaction-Typology Risk Scorer: OpenChainGraph compute node (risk_control). Regulatory deadline: 2027-07-01 (EU AMLR full application July 2027; AMLA full operations 2028). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: cry-01-zk-compliance-proof-generator, art-11-vop-batch-match-rate-analyser, ptg-01-ap2-prompt-template-generator, mms-03-app-fraud-graph, ml-01-isolation-forest. Open at: https://ainumbers.co/chaingraph/art-10-amla-transaction-typology-risk-scorer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Scores a synthetic transaction graph against AML typologies and FATF Travel Rule predicates; exports an AML risk determination per account/cluster. Outputs: average_score, compliance_flags, high_risk_count, max_score, medium_risk_count, overall_risk, top_risk_accounts, transaction_count, travel_rule_violations, typology_hit_counts. Output schema: call describe_tool("score_aml_typologies").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_cash_forecast_accuracyCash Forecast Accuracy Scoring: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Cash Forecast Accuracy Scoring: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-258-parse-camt053-reconciliation, art-261-test-hedge-effectiveness. Open at: https://ainumbers.co/chaingraph/art-263-score-cash-forecast-accuracy.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Scores treasury cash forecast accuracy per AFP Cash Forecasting Survey 2024 benchmarks. Outputs: afp_benchmark_tiers, by_horizon, not_legal_advice, overall_accuracy_tier, overall_bias_pct, overall_mape_pct, persistent_sign_periods, pii_note, regulatory_basis, skipped_zero_actual, table_source, table_version, timing_bias_detected, total_observations. Output schema: call describe_tool("score_cash_forecast_accuracy").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_clause_coverageClause Coverage Scorer: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Clause Coverage Scorer: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-410-clause-coverage-scorer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_credit_model_quantizedQuantized Credit Model Scorer: OpenChainGraph compute node (credit_assessment).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Quantized Credit Model Scorer: OpenChainGraph compute node (credit_assessment). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-348-score-credit-model-quantized.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Runs a fixed, int8-quantized logistic-regression-class credit-decisioning model as a pure integer inference kernel and returns the score it produced from the supplied normalized inputs. Outputs: accumulator_fixp, bits, decision, n_features, quant_method, threshold_fixp. Output schema: call describe_tool("score_credit_model_quantized").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_eudr_country_riskEUDR Country Benchmark Risk Scorer: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
EUDR Country Benchmark Risk Scorer: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-169-eudr-supply-chain-traceability-linker. Open at: https://ainumbers.co/chaingraph/art-168-eudr-country-benchmark-risk-scorer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Score country-of-production against the EUDR benchmark risk classification (low/standard/high per Art. Outputs: benchmark_risk, commission_delegated_act_note, country_code, due_diligence_level, inspection_rate_pct. Output schema: call describe_tool("score_eudr_country_risk").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_fuzzy_match_calibrationFuzzy-Match Calibration Scorer: OpenChainGraph compute node (model_governance).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Fuzzy-Match Calibration Scorer: OpenChainGraph compute node (model_governance). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-90-sanctions-screening-fit-diagnostic. Output feeds: art-97-sanctions-screening-quality-scorer, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-93-fuzzy-match-calibration-scorer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_mcp_readinessCompute a composite MCP server ship-readiness score across tool definitions, server.json, OAuth, transport, tool poisoning, and spec compliance.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Compute a composite MCP server ship-readiness score across tool definitions, server.json, OAuth, transport, tool poisoning, and spec compliance. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("score_mcp_readiness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_mcp_server_readinessMCP Developer Readiness Scorecard: OpenChainGraph compute node (compliance_control).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
MCP Developer Readiness Scorecard: OpenChainGraph compute node (compliance_control). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-17-ap2-mcp-policy-validator, art-23-visa-trusted-agent-protocol-inspector, art-24-mastercard-agentic-token-builder, art-25-a2a-agent-card-validator, art-26-x402-payload-decoder-flow-simulator, art-27-agentic-readiness-diagnostic, art-28-mcp-server-deployability-diagnostic. Open at: https://ainumbers.co/chaingraph/art-18-mcp-developer-readiness-scorecard.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Rolls up caller-supplied yes/partial/no answers across six MCP ship-readiness sections (tool definitions, server.json/registry, OAuth 2.1, transport security, tool-poisoning hygiene, spec-revision compliance) into an overall 0-100 score and a prioritized gap list. Outputs: answers_used, gaps, gaps_count, note, overall, sections, verdict. Output schema: call describe_tool("score_mcp_server_readiness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_mt_mx_translation_fidelityMT103 to MX Translation Fidelity Scorer: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
MT103 to MX Translation Fidelity Scorer: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-244-gpi-tracker-lifecycle-simulator. Open at: https://ainumbers.co/chaingraph/art-245-mt-mx-translation-fidelity-scorer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Scores ISO 15022 MT103 to ISO 20022 pacs.008 translation fidelity for CBPR+ migration. Outputs: cbpr_plus_deadline, charge_bearer_map, compliant, error_count, fidelity_score, fidelity_tier, issues, mapping_results, pii_note, regulatory_basis, table_source, table_version, truncation_risks. Output schema: call describe_tool("score_mt_mx_translation_fidelity").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_nis2_incident_significanceNIS2 Incident Significance Scorer (Art.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
NIS2 Incident Significance Scorer (Art. 23 Reporting Threshold): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-145-nis2-ict-supply-chain-diligence-scorer. Open at: https://ainumbers.co/chaingraph/art-144-nis2-incident-significance-scorer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Score whether an operational event meets the NIS2 Art. Outputs: early_warning_deadline_hours, estimated_affected_users, estimated_financial_loss_eur, final_report_deadline_days, notification_deadline_hours, recipients, reporting_required, service_disruption_hours, significance_verdict, triggering_factors. Output schema: call describe_tool("score_nis2_incident_significance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_nis2_supply_chain_diligenceNIS2 ICT Supply-Chain Diligence Scorer (Art.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
NIS2 ICT Supply-Chain Diligence Scorer (Art. 21(2)(d) / ENISA): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-144-nis2-incident-significance-scorer. Output feeds: art-146-nis2-governance-readiness-checker. Open at: https://ainumbers.co/chaingraph/art-145-nis2-ict-supply-chain-diligence-scorer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Score ICT vendor due-diligence posture against NIS2 Art. Outputs: active_risk_flags, enisa_control_coverage_pct, remediation_checklist, risk_score, risk_tier, vendor_incident_history_12mo. Output schema: call describe_tool("score_nis2_supply_chain_diligence").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_partner_stablecoin_readinessArc Partner Stablecoin Onboarding Conformance: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Arc Partner Stablecoin Onboarding Conformance: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-42-arc-fit-diagnostic. Output feeds: art-45-arc-xreserve-linter. Open at: https://ainumbers.co/chaingraph/art-110-arc-partner-stablecoin-onboarding.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Score a non-USD stablecoin issuer readiness to join Circle Partner Stablecoins on Arc against technical/operational, reserve-management, and risk-management standards. Outputs: ccy, composite_grade, eligible, gaps, grade, home_regime, reserve_score, risk_score, tech_score, verdict. Output schema: call describe_tool("score_partner_stablecoin_readiness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_payee_name_matchPayee Name-Match Score (VoP/CoP): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Payee Name-Match Score (VoP/CoP): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-376-score-payee-name-match.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Deterministic, versioned single-pair payee name-matching score for Verification-of-Payee / Confirmation-of-Payee evidence. Outputs: algorithm_version, close_match_threshold, compliance_flags, entity_suffix_stripped, match_band, match_threshold, normalized_account_name, normalized_reference_name, score, transliteration_in_scope. Output schema: call describe_tool("score_payee_name_match").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_sanctions_screening_qualitySanctions Screening-Program Quality Scorer: OpenChainGraph compute node (model_governance).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Sanctions Screening-Program Quality Scorer: OpenChainGraph compute node (model_governance). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-92-screening-list-coverage-checker, art-93-fuzzy-match-calibration-scorer. Output feeds: cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-97-sanctions-screening-quality-scorer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Wolfsberg-aligned screening-program quality scorecard: list coverage + match calibration + alert tuning + escalation workflow + model validation -> composite program-conformance grade and improvement priorities. Outputs: component_scores, component_weights, composite_pct, improvement_priorities, note, program_grade, reference_version, wolfsberg_note. Output schema: call describe_tool("score_sanctions_screening_quality").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_taxonomy_alignmentEU Taxonomy Alignment Scorer: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
EU Taxonomy Alignment Scorer: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-68-carbon-compliance-fit-diagnostic. Output feeds: art-74-taxonomy-kpi-gar-aggregator, art-75-eugb-factsheet-validator, cry-05-agent-action-audit-trail-aggregator. Open at: https://ainumbers.co/chaingraph/art-73-taxonomy-alignment-scorer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Scores an economic activity against an environmental objective: substantial-contribution technical-screening criteria + DNSH across the other five objectives + minimum safeguards -> aligned / eligible-but-not-aligned / not-eligible. Outputs: alignment_verdict, criterion_refs, dnsh_gaps, dnsh_results, is_aligned, nace_code, note, primary_objective, reference, safeguards_status, substantial_contribution_status. Output schema: call describe_tool("score_taxonomy_alignment").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
score_tempo_validator_readinessTempo Validator Readiness Scorer: OpenChainGraph compute node (infrastructure_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Tempo Validator Readiness Scorer: OpenChainGraph compute node (infrastructure_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-41-tempo-validator-readiness.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
screen_sanctions_privatePrivate-Input Sanctions Screen: OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Private-Input Sanctions Screen: OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-413-screen-sanctions-private.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Screens a privately held party/transfer list against a pinned OFAC-SDN-style list version and emits a public verdict (screened, hit count, clean flag) without disclosing the party list. Outputs: clean, coverage, hit_count, list_version, not_legal_advice, pii_note, regulatory_basis, screened. Output schema: call describe_tool("screen_sanctions_private").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
sdjwt_issueIssues a Selective Disclosure JWT (RFC 9901) from a claims object, marking selected top-level claim keys as selectively disclosable via salted-hash digests in the _sd array.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Description, as the server wrote it
Issues a Selective Disclosure JWT (RFC 9901) from a claims object, marking selected top-level claim keys as selectively disclosable via salted-hash digests in the _sd array. Signed EdDSA over a fresh ephemeral did:key (generated per call, not reused). Pair with sdjwt_present to build a redacted presentation from the returned sd_jwt.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: false
- openWorldHint: false
sdjwt_presentBuilds a redacted presentation from an existing SD-JWT (as returned by sdjwt_issue), keeping only the listed disclosures.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Description, as the server wrote it
Builds a redacted presentation from an existing SD-JWT (as returned by sdjwt_issue), keeping only the listed disclosures. Pass aud to add a KB-JWT holder-binding (a fresh ephemeral holder key is generated per call). Pass issuer_did (the sd_jwt's "issuer" field) to also verify the JWS and return the verifier_view -- exactly the claim set a relying party would resolve -- plus an OCG receipt of the presentation activity.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: false
- openWorldHint: false
select_agentic_checkout_protocolAgentic Checkout Protocol Selector: OpenChainGraph compute node (routing_policy).1 finding
Why it is in this group
- The name "select_agentic_checkout_protocol" says "select": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "select_agentic_checkout_protocol" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "select_agentic_checkout_protocol" says "select": a word for a change.
  What this check looks for
Description, as the server wrote it
Agentic Checkout Protocol Selector: OpenChainGraph compute node (routing_policy). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-20-acp-ucp-product-feed-conformance-auditor, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-19-agentic-checkout-protocol-selector.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Scores ACP, UCP, x402, and Visa TAP against platform profile (buyer type, AOV, geography, stack capabilities) and returns a ranked protocol recommendation with fit scores. Outputs: primary_name, primary_recommendation, primary_score, profile, protocol_scores, recommended_protocols, viable_protocols. Output schema: call describe_tool("select_agentic_checkout_protocol").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
select_cbe_licenseCan't Be Evil License Selector: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "select_cbe_license" says "select": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "select_cbe_license" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "select_cbe_license" says "select": a word for a change.
  What this check looks for
Description, as the server wrote it
Can't Be Evil License Selector: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-196-cant-be-evil-license-selector.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Rights-matrix lookup for a16z's six Can't Be Evil NFT licenses. Outputs: arweave_uri, arweave_uri_legacy, caveats, cbe_id, commercial, creator_retains, current_enum_name, derivatives, disclaimer, exclusivity, launch_alias, license_version_index, matrix_verified, objectionable_use_restriction, reference_url, sublicense. Output schema: call describe_tool("select_cbe_license").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
select_embedded_licenseEmbedded License Selector: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "select_embedded_license" says "select": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "select_embedded_license" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "select_embedded_license" says "select": a word for a change.
  What this check looks for
Description, as the server wrote it
Embedded License Selector: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-198-cross-license-rights-comparator. Output feeds: art-204-license-compatibility-checker. Open at: https://ainumbers.co/chaingraph/art-203-embedded-license-selector.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Maps creator answers to the SolSea and ALL.ART 4-tier embedded-license menu: Private/No Commercial, Personal/Public Display/No Commercial, Public Display/No Commercial, or Reproduction/Commercial. Outputs: decision_path, description, disclaimer, inputs_resolved, label, rights, source_family, source_url, tier_id. Output schema: call describe_tool("select_embedded_license").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
servicing_compliance_clockServicing Compliance Clock: OpenChainGraph compute node (consumer_finance).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Servicing Compliance Clock: OpenChainGraph compute node (consumer_finance). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/733-servicing-compliance-clock.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Computes servicer-compliance due dates and clock states from event dates and loan values across the Regulation X/Z servicing rules: error-resolution, information-request, loss-mitigation, foreclosure-timeline, force-placed-insurance, reminder, cancel, refund, payment-crediting, payoff-statement a... Outputs: verdict, obligations, evidence, manual_review_required. Output schema: call describe_tool("servicing_compliance_clock").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
simulate_agent_spend_policyAgentic Mandate Sandbox: OpenChainGraph compute node (agent_guardrail_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Agentic Mandate Sandbox: OpenChainGraph compute node (agent_guardrail_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-27-agentic-readiness-diagnostic. Output feeds: art-16-google-ap2-mandate-builder. Open at: https://ainumbers.co/chaingraph/art-15-agentic-mandate-sandbox.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Builds a deterministic Agent Guardrail Mandate skeleton from declared spend caps, MCC allowlist/blocklist, velocity rules, time windows, and approval thresholds. Outputs: approval_thresholds, compliance, corridor, mandate_id, mcc_constraints, note, rail, rejected_inputs, schema, spend_caps, time_windows, velocity_rules. Output schema: call describe_tool("simulate_agent_spend_policy").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
simulate_gpi_tracker_lifecycleSWIFT GPI Tracker Lifecycle Simulator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
SWIFT GPI Tracker Lifecycle Simulator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-245-mt-mx-translation-fidelity-scorer. Open at: https://ainumbers.co/chaingraph/art-244-gpi-tracker-lifecycle-simulator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates SWIFT GPI pacs.002 payment status code transitions against the GPI state machine (PDNG, ACSP, ACSP/ACWC, ACCC, RJCT) and checks the Universal Confirmation SLA (ACCC must be sent within 24 hours of ACSP). Outputs: allowed_next_statuses, amount_usd, current_status, hours_elapsed, is_rejected, is_settled, is_terminal, issues, lifecycle_states, next_status, pii_note, regulatory_basis, sla_breached, sla_hours_limit, sla_note, stage_description, table_source, table_version, transition_reason, transition_valid. Output schema: call describe_tool("simulate_gpi_tracker_lifecycle").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
simulate_output_floorBasel Output-Floor Phase-In Simulator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Basel Output-Floor Phase-In Simulator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-358-simulate-output-floor.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Basel III finalization / 2026 reproposal output-floor simulator: applies the published floor mechanic (applied RWA = max(internal-model RWA, floor% x standardized RWA)) across a caller-supplied annual phase-in schedule, returning the year-by-year capital impact path, the first binding-floor year,... Outputs: binding_floor_year, capital_impact_path, floor_ever_binds, internal_model_rwa, max_incremental_rwa, note, regulatory_basis, rule_status, standardized_rwa. Output schema: call describe_tool("simulate_output_floor").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
simulate_spend_policyAgent Spend-Policy Simulator: OpenChainGraph compute node (payment_policy).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Agent Spend-Policy Simulator: OpenChainGraph compute node (payment_policy). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-01-ap2-mandate-chain-validator, art-04-agent-identity-attestation-checker, art-32-a2a-agent-card-trust-chain-validator. Output feeds: art-01-ap2-mandate-chain-validator, art-04-agent-identity-attestation-checker, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-02-agent-spend-policy-simulator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Simulates thousands of synthetic agent transactions against a user-authored spend policy (per-merchant caps, category allow/deny, velocity limits, cumulative ceilings). Outputs: bypass_paths_detected, compliance_flags, fail_count, fail_rate_pct, pass_count, top_fail_reasons, total_approved_spend, total_transactions, verdict. Output schema: call describe_tool("simulate_spend_policy").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
simulate_var_monte_carloPortfolio VaR — Monte Carlo (Integer PRNG): OpenChainGraph compute node (risk_control).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Portfolio VaR — Monte Carlo (Integer PRNG): OpenChainGraph compute node (risk_control). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: qfa-02-portfolio-var-engine, qfa-03-stress-test-engine. Open at: https://ainumbers.co/chaingraph/art-371-simulate-var-monte-carlo.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Monte Carlo portfolio Value-at-Risk and Expected Shortfall over a one-factor correlated-asset model. Outputs: compliance_flags, conf_level, correlation, draw_count, es_dollar_mm, holding_period, mc_es_pct, mc_var_pct, n_assets, n_paths, prng_algorithm, seed, var_dollar_mm, verdict. Output schema: call describe_tool("simulate_var_monte_carlo").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
simulate_vop_matchingVoP Batch Match-Rate Analyser: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
VoP Batch Match-Rate Analyser: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: rca-03-iso20022-address-migration-verifier, art-10-amla-transaction-typology-risk-scorer. Output feeds: ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-11-vop-batch-match-rate-analyser.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Batch IBAN-name matching: match/close-match/no-match classification, configurable strictness (exact/normalized/fuzzy), false-positive vs false-negative trade-off curves, per-corridor mismatch-rate distribution. Outputs: close_match, match, match_rate_pct, no_match, total_records. Output schema: call describe_tool("simulate_vop_matching").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
simulate_x402_flowx402 Header Decoder, Payload Linter & 402 Flow Simulator: OpenChainGraph compute node (compliance_control).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
x402 Header Decoder, Payload Linter & 402 Flow Simulator: OpenChainGraph compute node (compliance_control). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-22-agentic-payments-protocol-comparator, art-25-a2a-agent-card-validator. Output feeds: art-03-x402-settlement-modeler, art-18-mcp-developer-readiness-scorecard, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-26-x402-payload-decoder-flow-simulator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Decodes base64 PAYMENT-REQUIRED / PAYMENT-SIGNATURE / PAYMENT-RESPONSE headers, lints exact-scheme PaymentPayload (EIP-3009 style authorization fields), walks the HTTP-402 request/verify/settle flow, shows scheme×network matrix. Outputs: decoded_type, errors, findings, has_accepts, is_json, mode, network, passes, scheme, score, warnings. Output schema: call describe_tool("simulate_x402_flow").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
stress_test_ap_redemption_pathAP Concentration + Redemption-Path Stress: OpenChainGraph compute node (collateral_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
AP Concentration + Redemption-Path Stress: OpenChainGraph compute node (collateral_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-322-rhc-ap-redemption-stress.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Stress-tests the one-token-equals-one-share economic-exposure claim for Robinhood Chain stock tokens against actual redemption reachability. Outputs: ap_count, concentration_risk, issuer_credit_exposure_distinct, liquidity_flag, not_investment_advice, premium_discount_exposure, redemption_path, redemption_reachable_for_non_ap, structural_dependencies, verdict. Output schema: call describe_tool("stress_test_ap_redemption_path").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
suggest_tool_ideaBuilds a prefilled GitHub Issue-Forms URL for proposing a new AINumbers fintech tool, OpenChainGraph compute node, or verifiable workflow.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Builds a prefilled GitHub Issue-Forms URL for proposing a new AINumbers fintech tool, OpenChainGraph compute node, or verifiable workflow. Returns the URL only -- it never posts to GitHub on the caller's behalf. Companion to the site's mailto suggestion form (suggest.html); this is the agent-callable path.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: true
suite_howtoHow to chain AINumbers tools into audited workflows: call with NO recipe_id for the compact recipe index, or with one recipe_id for that workflow's full step-by-step recipe.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
How to chain AINumbers tools into audited workflows: call with NO recipe_id for the compact recipe index, or with one recipe_id for that workflow's full step-by-step recipe. Every recipe carries the suite conventions end to end: Policy Mandate exports, execution_hash verification, and receipts.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
sweep_fedwire_addressesFedwire Payment-File Address Sweep: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Fedwire Payment-File Address Sweep: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-349-fedwire-structured-address-linter. Open at: https://ainumbers.co/chaingraph/art-350-fedwire-address-sweep.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Batch-sweeps a Fedwire or CHIPS payment file (CSV, one record per row) through the Fedwire/CHIPS structured-address lint (lint_fedwire_structured_address, art-349) per record. Outputs: disambiguation, fedwire_chips_deadline, pii_note, regulatory_basis, rejection_risk_report, risk_score, table_source, table_version, fedwire_chips_deadline_status. Output schema: call describe_tool("sweep_fedwire_addresses").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
test_bifsg_bias_thresholdsBIFSG Insurance Proxy Bias Threshold Test (Colorado SB 21-169): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
BIFSG Insurance Proxy Bias Threshold Test (Colorado SB 21-169): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-240-assess-naic-ais-program-readiness. Open at: https://ainumbers.co/chaingraph/art-239-test-bifsg-bias-thresholds.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Tests BIFSG proxy bias thresholds under Colorado SB 21-169 / Reg. Outputs: anchor_instruction, attestation_deadline, attestation_year, bias_detected, do_now, marginal_effect_flag, marginal_effect_pct, marginal_effect_threshold_pp, model_type, p_value, p_value_significant, p_value_threshold, pii_note, premium_flag, premium_per_1000_above_avg_pct, premium_threshold_pct, regulatory_basis, remediation_required, table_source, table_version, test_context, test_result. Output schema: call describe_tool("test_bifsg_bias_thresholds").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
test_hedge_effectivenessHedge Effectiveness Test: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Hedge Effectiveness Test: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-263-score-cash-forecast-accuracy. Open at: https://ainumbers.co/chaingraph/art-261-test-hedge-effectiveness.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: ASC 815-20-35 retrospective hedge effectiveness test. Outputs: anchor_surface, asc815_80_125_band, cumulative_hedged_change, cumulative_hedging_change, dollar_offset_effective, dollar_offset_ratio, effectiveness_reason, effectiveness_standard, hedge_ratio, ifrs9_hedge_ratio_passes, is_effective, method_applied, not_legal_advice, observation_count, ols_alpha, ols_beta, pii_note, r_squared, regression_effective, regulatory_basis, table_source, table_version. Output schema: call describe_tool("test_hedge_effectiveness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
test_hoepa_high_costHOEPA High-Cost Mortgage Trigger Test: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
HOEPA High-Cost Mortgage Trigger Test: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-220-reg-z-threshold-lookup. Open at: https://ainumbers.co/chaingraph/art-234-test-hoepa-high-cost.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Tests all three HOEPA high-cost mortgage triggers per Reg Z §1026.32(a)(1): (i) APR trigger (APOR+6.5pp first-lien, APOR+8.5pp subordinate or dwelling below $50k); (ii) points-and-fees trigger (5% of loan or $1,380 floor, 2026, FR 2025-22773); (iii) prepayment penalty trigger (>36 months or >2% o... Outputs: apor_pct, apr_pct, apr_spread_pct, apr_threshold_basis, apr_threshold_pct, apr_trigger_met, consumes, fr_citation, has_prepayment_penalty, is_high_cost, is_small_dwelling, lien_type, loan_amount, note, points_and_fees, points_fees_floor, points_fees_limit, points_fees_limit_pct, points_fees_trigger_met, pp_pct_limit, pp_period_limit_months, prepayment_penalty_pct, prepayment_penalty_period_months, prepayment_penalty_trigger_met, regulatory_basis, table_version, triggers_fired, year. Output schema: call describe_tool("test_hoepa_high_cost").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
test_hpml_escrowHPML Definition and Escrow Requirement Test: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
HPML Definition and Escrow Requirement Test: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-220-reg-z-threshold-lookup. Open at: https://ainumbers.co/chaingraph/art-235-test-hpml-escrow.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
test_nav_error_materialityTest NAV-Error Materiality: OpenChainGraph compute node (attestation_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Test NAV-Error Materiality: OpenChainGraph compute node (attestation_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-373-recompute-fund-nav. Open at: https://ainumbers.co/chaingraph/art-374-test-nav-error-materiality.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Compares an erroneous NAV-per-share against a corrected NAV-per-share against a DECLARED materiality policy (the industry half-cent absolute and 1% relative conventions, plus the fund's own policy taken as an input) using fixed-point BigInt money math throughout. Outputs: affected_period, declared_policy, error, estimated_impact, fence, fund_id, industry_convention, materiality_verdict, not_proven, regulatory_framework, reprocessing_need_indicated, shares_outstanding, structural_error, valuation_date. Output schema: call describe_tool("test_nav_error_materiality").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
track_ifrs17_loss_component_rollforwardIFRS 17 Loss Component Roll-Forward Tracker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
IFRS 17 Loss Component Roll-Forward Tracker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-178-ifrs17-csm-rollforward-validator. Open at: https://ainumbers.co/chaingraph/art-448-ifrs17-loss-component-tracker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Tracks the IFRS 17 para 50 loss-component roll-forward across periods: opening balance, additional loss recognised on new onerous contracts, reversal from subsequent favourable experience (capped so it never reverses more than the available balance), other adjustments, and release to profit or lo... Outputs: additional_lc, closing_lc, fully_reversed, lc_valid, opening_lc, other_adj, pre_release, release_to_pnl, reversal_lc. Output schema: call describe_tool("track_ifrs17_loss_component_rollforward").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_openids_homeowners_recordopenIDS Homeowners Record Validator: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "validate_openids_homeowners_record" says "validate": a word for reading.
- The description's first verb is "open": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "validate_openids_homeowners_record" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "validate_openids_homeowners_record" says "validate": a word for reading.
  What this check looks for
Description, as the server wrote it
openIDS Homeowners Record Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-256-validate-openids-homeowners-record.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates homeowners insurance data records against the openIDS Homeowners Data Standard v1.0 (AAIS / Linux Foundation, November 2025) -- the first free open (Apache-2.0) insurance data standard. Outputs: error_count, errors, not_legal_advice, pii_fields_found, pii_note, record_valid, regulatory_basis, section_results, sections_present, sections_required, standard_version, table_source, warning_count, warnings. Output schema: call describe_tool("validate_openids_homeowners_record").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_openvex_statementOpenVEX Statement Validator: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "validate_openvex_statement" says "validate": a word for reading.
- The description's first verb is "open": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "validate_openvex_statement" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "validate_openvex_statement" says "validate": a word for reading.
  What this check looks for
Description, as the server wrote it
OpenVEX Statement Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-136-slsa-provenance-verifier. Open at: https://ainumbers.co/chaingraph/art-137-openvex-statement-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates an OpenVEX document: @context includes openvex.dev, every statement carries vulnerability, products[], status in [not_affected,affected,fixed,under_investigation], and not_affected statements must include justification. Outputs: context_ok, invalid_statements, statement_count, vex_valid. Output schema: call describe_tool("validate_openvex_statement").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
vc_issueComposes and signs a W3C Verifiable Credential 2.0 with an eddsa-jcs-2022 Data Integrity proof over a fresh ephemeral did:key (generated per call, not reused).3 findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Findings
- Info
  The title mentions "Credential".
  Matched text
  title: "Issue a signed W3C Verifiable Credential 2.0"
  What this check looks for
- Info
  The description mentions "Credential".
  Matched text
  description: "Composes and signs a W3C Verifiable Credential 2.0 with an eddsa-jcs-2022 Data Integri..."
  What this check looks for
- Info
  The input schema at properties.subject_id.description mentions "Credential".
  Matched text
  inputSchema.properties.subject_id.description: "Credential subject id (DID or any identifier strin..."
  What this check looks for
Description, as the server wrote it
Composes and signs a W3C Verifiable Credential 2.0 with an eddsa-jcs-2022 Data Integrity proof over a fresh ephemeral did:key (generated per call, not reused). Returns the signed credential, an OCG Standard §23 vc-2.0 input-attestation block ready to embed in a ChainGraph chain's policy_parameters at the given pointer, and an OCG receipt of the issuance activity. The signature proves the claims were not altered after signing and that the did:key holder produced it -- it is not, by itself, identity assurance.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: false
- openWorldHint: false
verify_a2a_agent_cardA2A Agent Card Validator & Extension Checker: OpenChainGraph compute node (compliance_control).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
A2A Agent Card Validator & Extension Checker: OpenChainGraph compute node (compliance_control). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-22-agentic-payments-protocol-comparator. Output feeds: art-18-mcp-developer-readiness-scorecard, art-26-x402-payload-decoder-flow-simulator, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-25-a2a-agent-card-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates an A2A agent-card.json against the v1.0 shape: identity fields, capabilities, extensions (AP2/x402), input/output modes, skills, provider, signed-card JWS block. Outputs: errors, findings, has_ap2_extension, has_signed_card, passes, score, verdict, warnings. Output schema: call describe_tool("verify_a2a_agent_card").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_acdc_delegation_chainACDC Delegation Chain Verifier: OpenChainGraph compute node (compliance_mandate).2 findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Info
  The description mentions "credential".
  Matched text
  description: "...tic Chained Data Containers (ACDC): per-credential SAID self-addressing integrity, issuer-..."
  What this check looks for
- Info
  The input schema at properties.policy_parameters.description mentions "credentials".
  Matched text
  inputSchema.properties.policy_parameters.description: "...ompute is "auto" or "server". Required: credentials, max_chain_depth. Unknown keys are refu..."
  What this check looks for
Description, as the server wrote it
ACDC Delegation Chain Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-284-did-webvh-log-verifier. Open at: https://ainumbers.co/chaingraph/art-285-acdc-delegation-chain-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Verifies a chain of Authentic Chained Data Containers (ACDC): per-credential SAID self-addressing integrity, issuer-to-issuee edge linkage between successive credentials, schema SAID match, termination at a stated root AID, and revocation-status passthrough (report, never resolve). Outputs: chain_depth, edge_failures, revocation_status_reported, root_aid_matched, said_failures, valid. Output schema: call describe_tool("verify_acdc_delegation_chain").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_anchored_extractAnchored Extract Verifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Anchored Extract Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-286-anchored-extract-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Verifies an extract's Merkle inclusion against a root only when that root is anchored by a recognized source: a recognized OCG artifact/chain envelope, or an external anchor (RFC 3161 timestamp, OpenTimestamps, Sigstore transparency log, or an on-chain commitment composed via verify_eth_state_pro... Outputs: anchor_note, anchored, anchored_extract_determination, claimed_root, computed_root, escalation, not_proven, regulatory_framework, root_match, source_class, structural_error. Output schema: call describe_tool("verify_anchored_extract").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_content_credential_signatureContent Credential Signature Verifier: OpenChainGraph compute node (compliance_mandate).2 findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Info
  The title mentions "Credential".
  Matched text
  title: "Content Credential Signature Verifier"
  What this check looks for
- Info
  The description mentions "Credential".
  Matched text
  description: "Content Credential Signature Verifier: OpenChainGraph comp..."
  What this check looks for
Description, as the server wrote it
Content Credential Signature Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-123-c2pa-manifest-validator. Output feeds: art-125-provenance-ingredient-tree-resolver. Open at: https://ainumbers.co/chaingraph/art-124-content-credential-signature-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_conversion_receiptConversion Receipt Verifier: OpenChainGraph compute node (cryptographic_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Conversion Receipt Verifier: OpenChainGraph compute node (cryptographic_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-191-conversion-receipt-builder. Open at: https://ainumbers.co/chaingraph/art-192-conversion-receipt-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Re-verifies a conversion receipt from art-191: recomputes binding_sha256 over the JCS-canonical receipt and compares, checks structure and hex fields, and optionally compares digests re-hashed from the actual files. Outputs: binding_ok, checks, digest_ok, verdict. Output schema: call describe_tool("verify_conversion_receipt").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_disclosure_inclusionProves (or refutes) that a {path,digest} pair was in a disclosure manifest's room (DATAROOM-1-BUILD-SPEC.md §DR-4) -- worker-side mirror of tools/547-disclosure-manifest-verifier.h…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Proves (or refutes) that a {path,digest} pair was in a disclosure manifest's room (DATAROOM-1-BUILD-SPEC.md §DR-4) -- worker-side mirror of tools/547-disclosure-manifest-verifier.html's single-file inclusion check. Recomputes the manifest's Merkle root over its entries[], builds the inclusion proof path for the target entry, and reapplies it to confirm it reduces to the claimed merkle_root. Absence is only provable against the exact manifest version passed in.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_dscsa_transaction_statementDSCSA Transaction Statement (T3) Verifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
DSCSA Transaction Statement (T3) Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-113-saleable-returns-verifier. Open at: https://ainumbers.co/chaingraph/art-112-dscsa-transaction-statement-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Verify the DSCSA T3 set (Transaction Information + History + Statement) completeness, validate the GS1 SGTIN, and map the EPCIS 2.0 event type. Outputs: epcis_event, identifier_valid, missing_elements, t3_complete, transaction_date. Output schema: call describe_tool("verify_dscsa_transaction_statement").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_dual_layer_disclosureDual-Layer Disclosure Verifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Dual-Layer Disclosure Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-126-ai-act-art50-marking-checker. Output feeds: art-128-content-binding-assertion-validator. Open at: https://ainumbers.co/chaingraph/art-127-dual-layer-disclosure-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_einvoice_vat_calcE-Invoice VAT Calculation Verifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
E-Invoice VAT Calculation Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-293-einvoice-format-validator. Output feeds: art-295-einvoice-jurisdiction-mandate-router. Open at: https://ainumbers.co/chaingraph/art-294-einvoice-vat-calc-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Recompute an e-invoice's line VAT, per-category tax subtotals, tax total, and grand total from its line items under a supplied rounding convention, then compare against the document-asserted amounts. Outputs: consistent, grand_total_computed, grand_total_delta, parse_error, rounding, subtotal_deltas, tax_total_computed, tax_total_delta. Output schema: call describe_tool("verify_einvoice_vat_calc").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_eth_state_proofState-Proof Verifier: OpenChainGraph compute node (cryptographic_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
State-Proof Verifier: OpenChainGraph compute node (cryptographic_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-279-state-proof-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Verifies an EIP-1186 (eth_getProof) account and storage Merkle-Patricia-Trie proof against a caller-supplied trusted state root using a pure-JavaScript keccak-256 plus RLP decoder and bounded trie walk. Outputs: account, account_exists, address, block_state_root, bounded_limits, diagnostic, errors, proof_nodes_consumed, receipt_statement, regulatory_note, storage_results, verdict. Output schema: call describe_tool("verify_eth_state_proof").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_execution_hashIndependently verify a ChainGraph artifact (ChainGraph Standard v0.1 §6).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Independently verify a ChainGraph artifact (ChainGraph Standard v0.1 §6). Recomputes SHA-256 over the canonical (sorted-key, whitespace-stripped) JSON of policy_parameters + output_payload and compares it to the claimed execution_hash. A match proves the artifact's stated inputs deterministically produce its stated outputs. Pass either a full artifact object, or policy_parameters + output_payload + claimed_hash. Inputs must be I-JSON (RFC 7493): integers beyond 2^53 must be passed as JSON strings, and NaN/Infinity are rejected. Values outside I-JSON have no stable canonical form, so this tool returns a structured -32602 error (error.data.reason "ijson_violation") instead of a hash. Pure client-safe compute -- no data is stored. Use this to verify artifacts from any vendor that conforms to the ChainGraph Standard. Also verifies composite chain artifacts from run_chain (composite_artifact, ChainGraph SPEC §21.2/§21.4): hash-excluded adjacent metadata such as decision_trail, step_compliance_flags, non_live_steps and escalation_record is dropped before recompute, and the response names which rule applied.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_license_electionLicense Election Verifier: OpenChainGraph compute node (cryptographic_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
License Election Verifier: OpenChainGraph compute node (cryptographic_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-199-license-election-certifier. Open at: https://ainumbers.co/chaingraph/art-200-license-election-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Verifies a certificate produced by the License Election Certifier by recomputing the SHA-256 terms_hash over the JCS-canonical election core and comparing it against the stored hash. Outputs: binding_ok, checks, verdict. Output schema: call describe_tool("verify_license_election").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_product_authenticityLuxury Goods Product Authenticity Verifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Luxury Goods Product Authenticity Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-116-product-lineage-builder. Open at: https://ainumbers.co/chaingraph/art-117-product-authenticity-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Verify that presented lineage hashes chain back to the claimed root and that ownership transfers are continuous. Outputs: authentic, chains_to_root, ownership_continuous, product_id. Output schema: call describe_tool("verify_product_authenticity").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_reserve_proofReserve Proof Verifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Reserve Proof Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-280-reserve-proof-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Verifies a Merkle-sum Proof-of-Reserves customer-inclusion proof (OKX, Binance, Gate, Kraken export formats, or a generic canonical shape) against a declared root, and checks a Chainlink PoR / NAVLink aggregator round for staleness and deviation. Outputs: computed_leaf_hash, computed_root, declared_root, exchange, inclusion_verified, not_proven, por_round, regulatory_framework, reserve_proof_determination, root_hash_match, storage_proof_composition, structural_error, sum_verified. Output schema: call describe_tool("verify_reserve_proof").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_saleable_returnDSCSA Saleable Returns Verifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
DSCSA Saleable Returns Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-112-dscsa-transaction-statement-verifier. Output feeds: art-114-suspect-product-quarantine. Open at: https://ainumbers.co/chaingraph/art-113-saleable-returns-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Match a returned unit SGTIN+lot to its original transaction hash (DSCSA §582(c)(4)(D)). Outputs: id_match, lot_match, match, reason, txn_anchored, verdict. Output schema: call describe_tool("verify_saleable_return").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_slsa_provenanceSLSA Provenance Verifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
SLSA Provenance Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-135-cyclonedx-sbom-validator. Output feeds: art-137-openvex-statement-validator. Open at: https://ainumbers.co/chaingraph/art-136-slsa-provenance-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Verifies an in-toto SLSA provenance statement: validates _type (in-toto.io/Statement) and predicateType (slsa.dev/provenance), checks subject SHA-256 digest against caller-supplied artifact_digest_sha256, asserts builder.id present via runDetails.builder.id or predicate.builder.id, reports claime... Outputs: builder_id_present, pred_ok, provenance_valid, slsa_build_level, subject_digest_match, type_ok. Output schema: call describe_tool("verify_slsa_provenance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_timestamp_attestationTimestamp Attestation Verifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Timestamp Attestation Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-121-document-integrity-anchor. Open at: https://ainumbers.co/chaingraph/art-122-timestamp-attestation-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Recompute the document integrity anchor, confirm the document hash and timestamp claim match and the algorithm is consistent. Outputs: algo_match, hash_match, ts_consistent, verified. Output schema: call describe_tool("verify_timestamp_attestation").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_trade_document_setTrade Document Provenance & Consistency Verifier: OpenChainGraph compute node (cryptographic_mandate).1 finding
Why it is in this group
- The name "verify_trade_document_set" says "set": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "verify_trade_document_set" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "verify_trade_document_set" says "set": a word for a change.
  What this check looks for
Description, as the server wrote it
Trade Document Provenance & Consistency Verifier: OpenChainGraph compute node (cryptographic_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-52-digital-trade-fit-diagnostic, art-54-digital-trade-rules-checker. Output feeds: cry-04-merkle-batch-verifier, art-10-amla-transaction-typology-risk-scorer, ml-03-timeseries-anomaly-detector. Open at: https://ainumbers.co/chaingraph/art-55-trade-document-provenance-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Cross-validates a full trade-document set (eBL, commercial invoice, packing list, certificate of origin, insurance certificate) for internal consistency and computes a SHA-256 Merkle provenance root. Outputs: consistency_verdict, doc_count, invoicing_deviation_pct, merkle_root, mismatches, note, tbml_flags. Output schema: call describe_tool("verify_trade_document_set").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_trid_apr_accuracyTRID APR Accuracy Verifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
TRID APR Accuracy Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-215-reg-z-appendix-j-apr. Open at: https://ainumbers.co/chaingraph/art-217-trid-apr-accuracy.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: TRID APR accuracy check per Reg Z §1026.22(a). Outputs: abs_difference_pct, actual_apr_pct, difference_pct, disclosed_apr_pct, headroom_pct, irregularity_reasons, is_irregular_transaction, note, regulatory_basis, tolerance_pct, verdict, within_tolerance. Output schema: call describe_tool("verify_trid_apr_accuracy").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_webbotauth_signatureWeb Bot Auth Signature Verifier (RFC 9421): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Web Bot Auth Signature Verifier (RFC 9421): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-130-signature-directory-validator. Open at: https://ainumbers.co/chaingraph/art-129-webbotauth-signature-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_witness_cosignaturesWitness Cosignature Verifier: OpenChainGraph compute node (cryptographic_mandate).No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Witness Cosignature Verifier: OpenChainGraph compute node (cryptographic_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-424-witness-cosignature-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Verifies a C2SP tlog-checkpoint + witness-cosignature note (SPEC.md §20.2) offline: confirms the note's origin and root match a §20/§20.1 batch anchor's declared anchored_hash, then checks each pinned witness's cosignature against a k-of-n threshold. Outputs: __async, anchored_hash, checks, log_origin, mode, not_proven, parsed, threshold, witness_keys. Output schema: call describe_tool("verify_witness_cosignatures").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
workbook_csv_parseParses CSV text with the SAME strict RFC 4180 parser as the tools/554 browser workbook -- quoted fields, embedded quotes/commas/newlines, CRLF -- and REJECTS malformed input rather…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Parses CSV text with the SAME strict RFC 4180 parser as the tools/554 browser workbook -- quoted fields, embedded quotes/commas/newlines, CRLF -- and REJECTS malformed input rather than repairing it (deterministic beats forgiving for hash-stable digests). Numeric-looking and TRUE/FALSE cells are type-coerced; everything else stays a string. Returns the resulting A1-style cells (raw + evaluated value) plus the sheet's row/column extent, ready for workbook_evaluate or workbook_range_digest -- unless as_artifact is true, in which case it returns a full OCG v0.4 artifact instead (see as_artifact).
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
workbook_evaluateFeeds a set of A1-style cells (formulas and/or literals) through the SAME headless evaluator the tools/554 browser workbook uses -- topo-sorted dependency graph, ~20 functions (SUM…1 finding
Why it is in this group
- The name "workbook_evaluate" says "evaluate": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "workbook_evaluate" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "workbook_evaluate" says "evaluate": a word for a change.
  What this check looks for
Description, as the server wrote it
Feeds a set of A1-style cells (formulas and/or literals) through the SAME headless evaluator the tools/554 browser workbook uses -- topo-sorted dependency graph, ~20 functions (SUM AVG MIN MAX COUNT COUNTIF IF AND OR NOT ROUND ABS CONCAT LEN LEFT RIGHT TRIM UPPER LOWER SUMIF), cycles resolve to "#CYCLE!" and any non-finite result to "#NUM!" rather than propagating. Returns the computed value of every supplied cell -- pure, no digest, no receipt -- unless as_artifact is true, in which case it returns a full OCG v0.4 artifact instead (see as_artifact).
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
workbook_range_digestSpreadsheet-range digest only; NOT the general artifact emitter (use emit_chaingraph_artifact to wrap arbitrary computation output — this tool's as_artifact mode covers whole-sheet…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Spreadsheet-range digest only; NOT the general artifact emitter (use emit_chaingraph_artifact to wrap arbitrary computation output — this tool's as_artifact mode covers whole-sheet workbook runs only). Evaluates the supplied cells (same evaluator as workbook_evaluate) and computes a canonical values_digest over one A1-style range (e.g. "B2:D9") -- the exact `executionHash(values, {})` path the rest of OCG uses, per chaingraph/workbook/INPUT-MANIFEST.md (WB-2). Returns one `ranges[]` fragment of the Spreadsheet Input Manifest schema; assemble the full manifest (source.csv_digest, produced_by/produced_at) around it. Same range digested from the same cells in the tools/554 browser workbook always produces the same values_digest -- unless as_artifact is true, in which case it returns a full OCG v0.4 artifact over the WHOLE sheet instead of a single-range fragment (see as_artifact).
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false

Only reads 96Looks things up and changes nothing.Claude Code: allow

analyze_prediction_marketPrediction Market Analyzer: OpenChainGraph compute node (event_market_pnl).No findings
Why it is in this group
- The name "analyze_prediction_market" says "analyze": a word for reading.
Description, as the server wrote it
Prediction Market Analyzer: OpenChainGraph compute node (event_market_pnl). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-212-prediction-market-arbitrage. Open at: https://ainumbers.co/chaingraph/art-211-prediction-market-analyzer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Analyzes a binary or scalar prediction market position. Outputs: break_even_probability, brier_score, disclaimer, entry_price, expected_value, fee_paid, half_kelly_stake, implied_probability, kelly_fraction, log_score, mode, n_contracts, no_vig_fair_value, odds_american, odds_decimal, odds_fractional, payout, pnl, settlement_clamped, settlement_delta, settlement_in_range, settlement_value, side, strike, unit_value, venue, won. Output schema: call describe_tool("analyze_prediction_market").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
assess_suspect_product_statusDSCSA Suspect/Illegitimate Product Quarantine Assessor: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "assess_suspect_product_status" says "status": a word for reading.
Description, as the server wrote it
DSCSA Suspect/Illegitimate Product Quarantine Assessor: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-113-saleable-returns-verifier. Open at: https://ainumbers.co/chaingraph/art-114-suspect-product-quarantine.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Determine suspect vs illegitimate product status and required actions (quarantine, investigate, 72-hour FDA Form 3911 notification, trading partner notification). Outputs: required_actions, status. Output schema: call describe_tool("assess_suspect_product_status").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
build_ai_decision_log_recordAI Decision Log Record Builder (EU AI Act Art 12): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "build_ai_decision_log_record" says "log": a word for reading.
Description, as the server wrote it
AI Decision Log Record Builder (EU AI Act Art 12): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-237-validate-agent-audit-trail, art-238-classify-annex3-decisioning-obligations. Open at: https://ainumbers.co/chaingraph/art-236-build-ai-decision-log-record.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Builds an EU AI Act Art 12(2)-conformant decision log record for high-risk AI systems in financial services. Outputs: anchor_surface, anchor_tools, art12_completeness_score, art12_fields_present, chain_position, confidence, decision_label, enforcement_dates, input_digest, missing_art12_fields, model_id, model_version, operator_id, output_digest, override_by, override_flag, pii_note, record_status, regulatory_basis, retention_months, sha256_prev_record, subject_ref, system_context, table_version. Output schema: call describe_tool("build_ai_decision_log_record").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
check_screening_list_coverageScreening List-Coverage Checker: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "check_screening_list_coverage" says "list": a word for reading.
Description, as the server wrote it
Screening List-Coverage Checker: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-90-sanctions-screening-fit-diagnostic, art-91-ownership-50pct-aggregator. Output feeds: art-97-sanctions-screening-quality-scorer, cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-92-screening-list-coverage-checker.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Conformance check: validates a screening config against the required-coverage matrix for EU consolidated + UN + UK Sanctions List (post-OFSI-closure 28 Jan 2026) + OFAC SDN with correct jurisdictional-nexus gating. Outputs: coverage_grade, coverage_pct, missing_lists, nexus_gaps, note, ofsi_migration_note, reference_version, refresh_adequate, refresh_frequency_assessed, required_lists, sectoral_lists_screened. Output schema: call describe_tool("check_screening_list_coverage").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
checklist_validate_definitionValidates a checklist or SOP definition JSON against the CHECKRUN-1 schema (definition_id, title, semver version, non-empty steps[] each with step_id/title/instruction/evidence_req…No findings
Why it is in this group
- The name "checklist_validate_definition" says "validate": a word for reading.
Description, as the server wrote it
Validates a checklist or SOP definition JSON against the CHECKRUN-1 schema (definition_id, title, semver version, non-empty steps[] each with step_id/title/instruction/evidence_requirement (none|text|file-digest|attestation)/gate (blocking|advisory)). Returns valid:true/false plus a field-by-field error list. Pair with checklist_step_receipt to run the definition headlessly.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compare_agentic_payment_protocolsCompare agentic payment protocols (AP2, ACP/Shared Payment Token, x402, Visa TAP, Mastercard Agent Pay) across credential, signing, scope, rail, identity, and audit dimensions; opt…1 finding
Why it is in this group
- The name "compare_agentic_payment_protocols" says "compare": a word for reading.
Findings
- Info
  The description mentions "credential".
  Matched text
  description: "...Visa TAP, Mastercard Agent Pay) across credential, signing, scope, rail, identity, and au..."
  What this check looks for
Description, as the server wrote it
Compare agentic payment protocols (AP2, ACP/Shared Payment Token, x402, Visa TAP, Mastercard Agent Pay) across credential, signing, scope, rail, identity, and audit dimensions; optionally recommend a fit for a scenario. Read-only local lookup and comparison — no state change, no keys held, nothing sent or registered anywhere. Use when a developer or strategist needs to orient across the fragmenting agentic-payments standards. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("compare_agentic_payment_protocols").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compare_agentic_rail_protocolsAgentic Payments Protocol Comparator: OpenChainGraph compute node (routing_policy).No findings
Why it is in this group
- The name "compare_agentic_rail_protocols" says "compare": a word for reading.
Description, as the server wrote it
Agentic Payments Protocol Comparator: OpenChainGraph compute node (routing_policy). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-16-google-ap2-mandate-builder, art-23-visa-trusted-agent-protocol-inspector, art-24-mastercard-agentic-token-builder, art-25-a2a-agent-card-validator, art-26-x402-payload-decoder-flow-simulator, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-22-agentic-payments-protocol-comparator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Compares AP2, ACP, x402, Visa TAP, Mastercard Agentic Token, and Tempo MPP (Machine Payments Protocol) across 8 dimensions (backer, artifact, signing, scope, rail, identity, audit, status) and 6 agentic scenarios. Outputs: crosswalk, note, protocol_names, protocols_compared, protocols_detail, recommendation, scenario. Output schema: call describe_tool("compare_agentic_rail_protocols").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compare_basel_2023_vs_2026Basel 2023-vs-2026 Capital-Delta Comparator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "compare_basel_2023_vs_2026" says "compare": a word for reading.
Description, as the server wrote it
Basel 2023-vs-2026 Capital-Delta Comparator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-357-basel-2023-vs-2026-capital-delta-comparator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Runs the same portfolio through the 2023 Basel III Endgame NPR risk-weight framework and the 2026 reproposal (2026-03-19, three NPRs) framework, then reports the RWA and minimum-capital delta -- the 'reproduce the $87.7bn relief story on OUR book' tool. Outputs: business_indicator, constants_version, credit_rwa_2023, credit_rwa_2026, delta_capital, delta_capital_pct, delta_rwa, direction, disambiguation, op_capital_2023, op_capital_2026, op_rwa_2023, op_rwa_2026, portfolio_summary, referenced_tool_ids, rule_status, source, total_capital_2023, total_capital_2026, total_rwa_2023, total_rwa_2026, unrecognized_asset_class. Output schema: call describe_tool("compare_basel_2023_vs_2026").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compare_corridor_costCorridor Cost Comparator (World Bank RPW): OpenChainGraph compute node (analytics_mandate).No findings
Why it is in this group
- The name "compare_corridor_cost" says "compare": a word for reading.
Description, as the server wrote it
Corridor Cost Comparator (World Bank RPW): OpenChainGraph compute node (analytics_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-248-compute-remittance-disclosure, art-250-model-stablecoin-corridor-economics. Open at: https://ainumbers.co/chaingraph/art-249-compare-corridor-cost.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Benchmarks a remittance corridor total cost (fee % + FX margin %) against the World Bank Remittance Prices Worldwide (RPW) Q1 2026 snapshot for the same origin-destination pair, the SmaRT global average (6.36%), and the SDG 10.c 3% target. Outputs: corridor, cost_at_200_usd, cost_at_500_usd, disambiguation, fee_pct, from_country, fx_margin_pct, meets_sdg_target, pii_note, regulatory_basis, rpw_benchmark_pct, rpw_corridor_avg_200, rpw_corridor_avg_500, rpw_service_count, sdg_target_pct, send_amount, service_name, smart_global_avg_pct, table_source, table_version, to_country, total_cost_pct, vs_rpw_benchmark, vs_sdg_target, vs_smart_avg. Output schema: call describe_tool("compare_corridor_cost").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compare_model_outcome_analysisModel Outcome-Analysis Comparison: OpenChainGraph compute node (compliance_control).No findings
Why it is in this group
- The name "compare_model_outcome_analysis" says "compare": a word for reading.
Description, as the server wrote it
Model Outcome-Analysis Comparison: OpenChainGraph compute node (compliance_control). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-450-model-inventory-entry. Output feeds: art-453-model-validation-status. Open at: https://ainumbers.co/chaingraph/art-451-model-outcome-analysis.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: SR 26-2 ongoing-monitoring backtest: compares a list of period predicted-vs-actual model outcomes, computes per-period absolute percent error, mean/max absolute percent error, and flags periods breaching a caller-declared error tolerance. Outputs: breach_periods, breach_rate_pct, error_threshold_pct, max_absolute_percent_error, max_breach_rate_pct, mean_absolute_percent_error, outcome_status, periods, total_periods, worst_period. Output schema: call describe_tool("compare_model_outcome_analysis").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compare_pension_lump_sum_annuityPension Lump-Sum vs.No findings
Why it is in this group
- The name "compare_pension_lump_sum_annuity" says "compare": a word for reading.
Description, as the server wrote it
Pension Lump-Sum vs. Annuity Decision Engine: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-282-social-security-claiming-optimizer. Open at: https://ainumbers.co/chaingraph/art-283-pension-lump-sum-vs-annuity-decision-engine.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Compares a defined-benefit pension lump-sum offer against the single-life and joint-survivor annuity streams: present value at the stated discount rate with an optional labeled COLA assumption, survivor-option monthly cost, break-even discount rate, and undiscounted break-even age. Outputs: annuityPV, breakEvenAge, breakEvenRate, colaApplied, lumpSum, recommendation, reinvestmentRequiredReturn, survivorOptionCostMonthly, survivorPct. Output schema: call describe_tool("compare_pension_lump_sum_annuity").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
compare_rights_matrixCross-License Rights Comparator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "compare_rights_matrix" says "compare": a word for reading.
Description, as the server wrote it
Cross-License Rights Comparator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-198-cross-license-rights-comparator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Compares any two licenses from the CC, CBE, and PIL families on a 9-dimension rights vector (copy, display, commercial, exclusive, modify, sublicense, share_alike, attribution, revocable). Outputs: checks, diff, dimensions, disclaimer, matrix_note, more_permissive_than, ref_a, ref_b, sources, vector_a, vector_b. Output schema: call describe_tool("compare_rights_matrix").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
describe_toolReturns the full definition of exactly one AINumbers MCP tool: name, description, inputSchema, outputSchema (the JSON Schema its structuredContent validates against), annotations, …No findings
Why it is in this group
- The name "describe_tool" says "describe": a word for reading.
Description, as the server wrote it
Returns the full definition of exactly one AINumbers MCP tool: name, description, inputSchema, outputSchema (the JSON Schema its structuredContent validates against), annotations, and lifecycle_status. Use it after find_tool or find_chain hands you an mcp_name, or after tools/list whose entries omit outputSchema to keep the catalog reply small — the pointer sentence "Output schema: call describe_tool(\"<name>\")." marks exactly those tools. One call, no side effects, zero network on the server: the definition comes from the vendored catalog. Unknown names are rejected with -32602 plus the nearest registered names.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
find_chainBM25 search over all 388 AINumbers ChainGraph chains.No findings
Why it is in this group
- The name "find_chain" says "find": a word for reading.
Description, as the server wrote it
BM25 search over all 388 AINumbers ChainGraph chains. Returns ranked chains with their full recipe: ordered node sequence, deep-links, composer URL, and entry tool mcp_name. Agent flow: find_chain(query) → read recipe → call the listed node MCP tools in order, passing parent_hashes between steps. Task-shaped queries work well: "reserve composition recompute", "FR 2052a classification", "CCP margin replication", "SR 26-2 model validation benchmark", "litigation damages recompute", "evidence pack assembly".
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
find_prediction_arbitragePrediction Market Arbitrage: OpenChainGraph compute node (event_market_pnl).No findings
Why it is in this group
- The name "find_prediction_arbitrage" says "find": a word for reading.
Description, as the server wrote it
Prediction Market Arbitrage: OpenChainGraph compute node (event_market_pnl). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-211-prediction-market-analyzer. Open at: https://ainumbers.co/chaingraph/art-212-prediction-market-arbitrage.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Identifies cross-venue arbitrage opportunities in binary prediction markets. Outputs: arb_exists, capital_deployed, consensus_gap, disclaimer, fee_a, fee_b, gross_profit, gross_spread, gross_spread_pct, implied_prob_yes_a, implied_prob_yes_b, k_contracts, min_spread_to_break_even, net_edge_pct, net_profit, no_price_b, payout, stake_total, total_fees, venue_a, venue_b, yes_price_a. Output schema: call describe_tool("find_prediction_arbitrage").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
find_toolBM25 search over all 685 live AINumbers ChainGraph node tools.No findings
Why it is in this group
- The name "find_tool" says "find": a word for reading.
Description, as the server wrote it
BM25 search over all 685 live AINumbers ChainGraph node tools. Returns ranked tools with mcp_name, URL, mandate type, and wave. Use to locate a specific computation node (e.g. "FRTB expected shortfall", "MiCA own funds", "XVA calculator", "reserve composition", "2052a", "initial margin", "model validation benchmark", "TVM / damages calculator") before calling it. Complements find_chain (chain-level) and list_ainumbers_tools (catalog-level).
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
ha_gate_statusGiven a step's haGatePolicy and the human_accountability_records[] collected for a subject, returns whether the gate is satisfied, held, rejected, escalated, or overridden -- same …No findings
Why it is in this group
- The name "ha_gate_status" says "status": a word for reading.
Description, as the server wrote it
Given a step's haGatePolicy and the human_accountability_records[] collected for a subject, returns whether the gate is satisfied, held, rejected, escalated, or overridden -- same algorithm as the browser verify.html HA gate card (kernels/_hagate.mjs, SPEC.md §27.4-27.5). A rejection record is terminal-blocking regardless of policy; an active time-boxed override (§27.5) takes precedence over the underlying policy. Pure function: caller supplies nowISO for determinism.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
ha_record_validateStructural (non-cryptographic) §27.2 signed-named-human check: does this record carry a §16 whole-artifact proof (eddsa-jcs-2022) whose verificationMethod is bound to the record's …No findings
Why it is in this group
- The name "ha_record_validate" says "validate": a word for reading.
Description, as the server wrote it
Structural (non-cryptographic) §27.2 signed-named-human check: does this record carry a §16 whole-artifact proof (eddsa-jcs-2022) whose verificationMethod is bound to the record's own identity.id? Does NOT verify the signature bytes -- pair with verify_execution_hash / a proof verifier for that. Use before counting a record toward ha_gate_status.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
inspect_visa_tap_signatureInspect a Visa Trusted Agent Protocol HTTP Message Signature and score TAP readiness.No findings
Why it is in this group
- The name "inspect_visa_tap_signature" says "inspect": a word for reading.
Description, as the server wrote it
Inspect a Visa Trusted Agent Protocol HTTP Message Signature and score TAP readiness. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("inspect_visa_tap_signature").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
inspect_visa_trusted_agent_protocolVisa Trusted Agent Protocol (TAP) Signature Inspector: OpenChainGraph compute node (compliance_control).No findings
Why it is in this group
- The name "inspect_visa_trusted_agent_protocol" says "inspect": a word for reading.
Description, as the server wrote it
Visa Trusted Agent Protocol (TAP) Signature Inspector: OpenChainGraph compute node (compliance_control). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-16-google-ap2-mandate-builder, art-22-agentic-payments-protocol-comparator. Output feeds: art-18-mcp-developer-readiness-scorecard, art-24-mastercard-agentic-token-builder, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-23-visa-trusted-agent-protocol-inspector.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Parses and scores Visa TAP Signature-Input / Signature header pairs (RFC 9421 HTTP Message Signatures). Outputs: errors, findings, parsed_label, parsed_params, passes, score, verdict, warnings. Output schema: call describe_tool("inspect_visa_trusted_agent_protocol").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lei_kyb_checkFetches an entity's live GLEIF LEI record (api.gleif.org -- the worker's only egress target for this tool, RULINGS-2026-07-19-EGRESS-APEX.md R1) and grades it across six GLEIF data…No findings
Why it is in this group
- The description's first verb is "fetch": a word for reading.
Description, as the server wrote it
Fetches an entity's live GLEIF LEI record (api.gleif.org -- the worker's only egress target for this tool, RULINGS-2026-07-19-EGRESS-APEX.md R1) and grades it across six GLEIF data-quality dimensions -- registration status, renewal timeliness, corroboration level, entity status, Level-2 parent disclosure, address completeness -- using the SAME deterministic grader as the tools/551 browser workbench (pasting the same record there reproduces identical grades). Returns per-dimension grades, a composite, and a receipt with source provenance (url, retrieved_at, response_digest). This is an ASSERTED FETCH, not a zkTLS proof of origin, and grades DATA QUALITY only -- never creditworthiness, sanctions status, or entity legitimacy.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: false
- openWorldHint: true
lint_mcp_tool_definitionValidate an MCP tool definition against JSON Schema 2020-12 and current naming, output-schema, and annotation rules; returns findings, a conformance score, and a recommended annota…No findings
Why it is in this group
- The description's first verb is "validate": a word for reading.
Description, as the server wrote it
Validate an MCP tool definition against JSON Schema 2020-12 and current naming, output-schema, and annotation rules; returns findings, a conformance score, and a recommended annotation set. Use when a developer wants to check an MCP tool definition before publishing. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("lint_mcp_tool_definition").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
list_ainumbers_toolsSearch the AINumbers catalog (1200+ client-side fintech tools).No findings
Why it is in this group
- The name "list_ainumbers_tools" says "list": a word for reading.
Description, as the server wrote it
Search the AINumbers catalog (1200+ client-side fintech tools). Returns deep-links; prefill-enabled tools accept #in=<base64url(JSON of {element_id: value})>[&run=1] for one-click invocation.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lookup_mletr_jurisdiction_adoptionMLETR Jurisdiction-Adoption Lookup: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "lookup_mletr_jurisdiction_adoption" says "lookup": a word for reading.
Description, as the server wrote it
MLETR Jurisdiction-Adoption Lookup: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-354-mletr-jurisdiction-adoption-lookup.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Static citation-table lookup of UNCITRAL MLETR (Model Law on Electronic Transferable Records) adoption status per jurisdiction -- statute, scope, and effective date, data_version stamped -- and a corridor verdict for whether an electronic bill of lading (eBL) is legally effective end-to-end betwe... Outputs: corridor, data_version, disambiguation, ebl_legally_effective, table_source, verdict. Output schema: call describe_tool("lookup_mletr_jurisdiction_adoption").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lookup_reg_z_thresholdsReg Z Threshold Lookup: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "lookup_reg_z_thresholds" says "lookup": a word for reading.
Description, as the server wrote it
Reg Z Threshold Lookup: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-220-reg-z-threshold-lookup.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Reg Z version-pinned threshold lookup service. Outputs: available_years, data, note, regulatory_basis, table, year. Output schema: call describe_tool("lookup_reg_z_thresholds").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
otlp_validateStructurally validates an OTLP/JSON trace (resourceSpans -> scopeSpans -> spans: hex traceId/spanId shape, quoted nanosecond timestamps, parent-span references, time ordering) and …No findings
Why it is in this group
- The name "otlp_validate" says "validate": a word for reading.
Description, as the server wrote it
Structurally validates an OTLP/JSON trace (resourceSpans -> scopeSpans -> spans: hex traceId/spanId shape, quoted nanosecond timestamps, parent-span references, time ordering) and checks every gen_ai span's attributes against a pinned gen_ai semantic-convention attribute snapshot (the upstream open-telemetry/semantic-conventions-genai namespace ships zero tagged releases, so this is a dated pin, stamped as semconv_snapshot in the result -- expect it to drift as the upstream spec moves). Byte-identical lint engine to tools/556-otlp-genai-span-composer-linter.html.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
pain001_validateValidates a pain.001.001.09 customer credit transfer initiation XML document against the same hand-derived schema-subset table as the tools/555 browser validator -- structural (man…No findings
Why it is in this group
- The name "pain001_validate" says "validate": a word for reading.
Description, as the server wrote it
Validates a pain.001.001.09 customer credit transfer initiation XML document against the same hand-derived schema-subset table as the tools/555 browser validator -- structural (mandatory fields), facet (IBAN mod-97, BIC format, ISO 4217 currency, decimal/date patterns), and batch-total cross-checks (NbOfTxs vs. actual transaction count, CtrlSum vs. sum of InstdAmt). Byte-identical error set to the browser tool for the same input. Subset validation, not full XSD conformance -- prepare/validate only, never transmits or generates a live payment message.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
redline_diffAn agent proposes an edit: diffs original text/markdown against its own revised version and returns the line-level hunks plus a diff receipt (original/revised digests, diff-algorit…No findings
Why it is in this group
- The name "redline_diff" says "diff": a word for reading.
Description, as the server wrote it
An agent proposes an edit: diffs original text/markdown against its own revised version and returns the line-level hunks plus a diff receipt (original/revised digests, diff-algorithm declaration, per-hunk digests). Byte-identical to what the tools/552 browser workbench computes for the same inputs -- a human can paste the SAME original/revised text into that workbench, disposition each hunk (accept/reject/comment), and their resulting hunk/disposition receipts will reference this diff_receipt's execution_hash. Pass the whole bundle (this diff_receipt + the human's hunk_receipts + disposition_receipt) to redline_verify to check the interleaved chain end-to-end.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
resolve_cbam_default_valueCBAM Default-Value Resolver: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "resolve_cbam_default_value" says "resolve": a word for reading.
Description, as the server wrote it
CBAM Default-Value Resolver: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-68-carbon-compliance-fit-diagnostic. Output feeds: art-69-cbam-embedded-emissions-calculator. Open at: https://ainumbers.co/chaingraph/art-70-cbam-default-value-resolver.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Resolves the Commission default embedded-emissions value for a (CN code x country-of-origin) pair, applies the year-dependent markup vs the actual-data path (+10% 2026, +20% 2027, +30% 2028+), and returns the value with provenance citing the CBAM Implementing Regulation. Outputs: actual_path_recommended, cn_code, country_of_origin, default_value_tco2e_per_t, effective_default, good_category, markup_pct, note, provenance, reference_version, reporting_year. Output schema: call describe_tool("resolve_cbam_default_value").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
resolve_provenance_ingredient_treeProvenance Ingredient Tree Resolver: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "resolve_provenance_ingredient_tree" says "resolve": a word for reading.
Findings
- Info
  The description mentions "credential".
  Matched text
  description: "...pstream artifacts from: art-124-content-credential-signature-verifier. Open at: https://ai..."
  What this check looks for
Description, as the server wrote it
Provenance Ingredient Tree Resolver: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-124-content-credential-signature-verifier. Open at: https://ainumbers.co/chaingraph/art-125-provenance-ingredient-tree-resolver.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
resolve_recall_traceFSMA 204 Recall Trace Resolver (24-Hour FDA List): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "resolve_recall_trace" says "resolve": a word for reading.
Description, as the server wrote it
FSMA 204 Recall Trace Resolver (24-Hour FDA List): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-119-traceability-lot-code-linker. Open at: https://ainumbers.co/chaingraph/art-120-recall-trace-resolver.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: One-up/one-back trace from a contaminated Traceability Lot Code to affected recipients and sources. Outputs: contaminated_tlc, recipients, sources, traced_count. Output schema: call describe_tool("resolve_recall_trace").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_a2a_agent_cardValidate an A2A agent-card.json against the v1.0 shape, check signatures, and confirm extension declarations.No findings
Why it is in this group
- The name "validate_a2a_agent_card" says "validate": a word for reading.
Description, as the server wrote it
Validate an A2A agent-card.json against the v1.0 shape, check signatures, and confirm extension declarations. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("validate_a2a_agent_card").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_a2a_trust_chainA2A Agent-Card Trust-Chain Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_a2a_trust_chain" says "validate": a word for reading.
Description, as the server wrote it
A2A Agent-Card Trust-Chain Validator: OpenChainGraph compute node (compliance_mandate). Regulatory deadline: 2026-08 (A2A at Linux Foundation (150+ orgs); EU AI Act dates re-set by the Digital Omnibus on AI, Regulation (EU) 2026/1744 (in force 27 Jul 2026): Chapter III high-risk regime (Sections 1-3) applies 2 Dec 2027 for Art. 6(2)/Annex III systems and 2 Aug 2028 for Art. 6(1)/Annex I systems; Art. 50(2) synthetic-content marking for systems placed on the market before 2 Aug 2026 applies by 2 Dec 2026 (amended Art. 111(4)); new Art. 5(1) points (ba)/(bb) prohibitions likewise apply from 2 Dec 2026. These push agent KYA toward requirement.). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-04-agent-identity-attestation-checker. Output feeds: art-04-agent-identity-attestation-checker, art-02-agent-spend-policy-simulator, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-32-a2a-agent-card-trust-chain-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: The horizontal agent-to-agent trust complement. Outputs: card_schema_ok, checks, fail_count, no_expired_links, no_scope_escalation, pass_count, signature_block_present, trust_determination, warn_count. Output schema: call describe_tool("validate_a2a_trust_chain").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_a2a_x402_mandateA2A x402-Extension Mandate Validator: OpenChainGraph compute node (settlement_mandate).No findings
Why it is in this group
- The name "validate_a2a_x402_mandate" says "validate": a word for reading.
Description, as the server wrote it
A2A x402-Extension Mandate Validator: OpenChainGraph compute node (settlement_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-03-x402-settlement-modeler. Output feeds: art-30-agent-commerce-conformance-validator, cry-05-agent-action-audit-trail-aggregator, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-31-a2a-x402-extension-mandate-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates the A2A x402 extension (Coinbase/MetaMask/Ethereum Foundation) that carries crypto-payment authority inside an AP2 mandate: extension declaration in the A2A agent card, payment-authority scope, settlement-rail binding, exact-scheme x402 PaymentPayload lint, and mandate-to-payment-leg co... Outputs: checks, extension_declared, fail_count, pass_count, payment_authority_scope_present, settlement_rail_bound, verdict, warn_count. Output schema: call describe_tool("validate_a2a_x402_mandate").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_acp_checkoutACP Checkout Conformance Validator: OpenChainGraph compute node (payment_mandate).No findings
Why it is in this group
- The name "validate_acp_checkout" says "validate": a word for reading.
Description, as the server wrote it
ACP Checkout Conformance Validator: OpenChainGraph compute node (payment_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-01-ap2-mandate-chain-validator, art-564-ucp-checkout-payload-lint. Output feeds: art-01-ap2-mandate-chain-validator, art-03-x402-settlement-modeler, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-12-acp-checkout-conformance-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: OpenAI/Stripe Agentic Commerce Protocol (ACP): CheckoutRequest/Response field conformance (10 required fields each), Shared Payment Token structure, ISO 4217 currency, TTL, signature prefix validation. Outputs: checks, currency, fail_count, merchant_id, overall_status, pass_count, warn_count. Output schema: call describe_tool("validate_acp_checkout").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_adverse_action_noticeValidate Adverse Action Notice: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_adverse_action_notice" says "validate": a word for reading.
Description, as the server wrote it
Validate Adverse Action Notice: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-228-build-adverse-action-notice. Open at: https://ainumbers.co/chaingraph/art-227-validate-adverse-action-notice.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates an adverse action notice against Reg B §1002.9 completeness requirements: reason count (max 4), prohibited vague reason codes (CFPB Circulars 2022-03/2023-03), required content fields (creditor name, date, statement of rights), and FCRA §615(a) rights disclosure when a credit score was... Outputs: action_taken, compliance_score, compliant, credit_score_used, fcra_required, fcra_violations, pii_note, reason_count, reason_count_valid, regulatory_basis, table_source, table_version, violation_count, violations, warning_count, warnings. Output schema: call describe_tool("validate_adverse_action_notice").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_agent_audit_trailAgent Audit Trail Conformance Validator (IETF AAT): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_agent_audit_trail" says "validate": a word for reading.
Description, as the server wrote it
Agent Audit Trail Conformance Validator (IETF AAT): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-236-build-ai-decision-log-record. Open at: https://ainumbers.co/chaingraph/art-237-validate-agent-audit-trail.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates agent audit trail records against IETF draft-sharif-agent-audit-trail-00 (AAT, expires Sept 2026). Outputs: aat_completeness_score, aat_required_fields_present, action_class, action_detail, agent_identity, alignment_note, chain_position, conformance_result, ecdsa_present, outcome, record_id, regulatory_basis, sha256_chain_format_valid, sha256_prev_record, table_version, trust_level, validation_errors, validation_warnings. Output schema: call describe_tool("validate_agent_audit_trail").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_agent_commerce_conformanceAgent Commerce Cross-Protocol Conformance Validator: OpenChainGraph compute node (payment_mandate).No findings
Why it is in this group
- The name "validate_agent_commerce_conformance" says "validate": a word for reading.
Description, as the server wrote it
Agent Commerce Cross-Protocol Conformance Validator: OpenChainGraph compute node (payment_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-01-ap2-mandate-chain-validator, art-03-x402-settlement-modeler, art-12-acp-checkout-conformance-validator, art-62-ap2-payment-receipt-verifier. Output feeds: cry-05-agent-action-audit-trail-aggregator, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-30-agent-commerce-conformance-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: The synergy flagship. Outputs: checks, fail_count, overall_status, pass_count, protocols_validated, warn_count. Output schema: call describe_tool("validate_agent_commerce_conformance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_agent_obo_mandateAgent On-Behalf-Of (OBO) Mandate Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_agent_obo_mandate" says "validate": a word for reading.
Description, as the server wrote it
Agent On-Behalf-Of (OBO) Mandate Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-150-mcp-tool-scope-revocation-auditor. Output feeds: art-152-mcp-task-lifecycle-validator. Open at: https://ainumbers.co/chaingraph/art-151-agent-obo-mandate-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate an agent on-behalf-of (OBO) mandate: subject (the user being represented), bounded scope array, intent string, and a non-expired validity window (caller-supplied now_unix, no clock reads). Outputs: gaps, has_intent, has_scope, has_subject, not_expired, verdict. Output schema: call describe_tool("validate_agent_obo_mandate").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_ai_impact_assessmentAI Risk Impact Assessment Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_ai_impact_assessment" says "validate": a word for reading.
Description, as the server wrote it
AI Risk Impact Assessment Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-171-iso42001-aims-clause-conformance. Output feeds: art-173-ai-system-governance-classifier. Open at: https://ainumbers.co/chaingraph/art-172-ai-risk-impact-assessment-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate ISO 42005-style AI impact-assessment completeness across seven required elements: intended use, affected stakeholders (>=1), risk treatment definition, monitoring plan, formal approval, risk category identification (>=1), and data source documentation. Outputs: complete, completeness_score, fields_checked, fields_passed, gaps. Output schema: call describe_tool("validate_ai_impact_assessment").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_ap2_mandate_chainAP2 Mandate-Chain Validator: OpenChainGraph compute node (payment_mandate).No findings
Why it is in this group
- The name "validate_ap2_mandate_chain" says "validate": a word for reading.
Description, as the server wrote it
AP2 Mandate-Chain Validator: OpenChainGraph compute node (payment_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-02-agent-spend-policy-simulator, art-12-acp-checkout-conformance-validator, art-36-tempo-mpp-agent-mandate. Output feeds: art-02-agent-spend-policy-simulator, art-03-x402-settlement-modeler, art-04-agent-identity-attestation-checker, art-12-acp-checkout-conformance-validator, art-385-agent-token-scope-checker, art-62-ap2-payment-receipt-verifier, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-01-ap2-mandate-chain-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates AP2 v0.2 Intent→Cart→Payment mandate trio: signature-chain integrity, scope/limit consistency, TTL/expiry, over-spend detection, Human-Not-Present autonomous-agent flows. Outputs: validation_verdict, checks_run, failing_checks, warning_checks, mandate_ids, has_cart, human_not_present. Output schema: call describe_tool("validate_ap2_mandate_chain").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_ap2_mandate_credentialAP2/MCP Policy Validator: OpenChainGraph compute node (scheme_rule).No findings
Why it is in this group
- The name "validate_ap2_mandate_credential" says "validate": a word for reading.
Description, as the server wrote it
AP2/MCP Policy Validator: OpenChainGraph compute node (scheme_rule). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-16-google-ap2-mandate-builder, art-27-agentic-readiness-diagnostic. Output feeds: art-18-mcp-developer-readiness-scorecard. Open at: https://ainumbers.co/chaingraph/art-17-ap2-mcp-policy-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_ap2_mcp_policyValidate AP2 Policy Mandate JSON payloads against the Unified Build Contract v1.0 schema.No findings
Why it is in this group
- The name "validate_ap2_mcp_policy" says "validate": a word for reading.
Description, as the server wrote it
Validate AP2 Policy Mandate JSON payloads against the Unified Build Contract v1.0 schema. Auto-generates MCP tool definitions from the mandate and simulates agent ingestion of agent_instructions. Use when authoring or testing AP2 agentic payment policies. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("validate_ap2_mcp_policy").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_c2pa_manifestC2PA Content Credential Manifest Validator: OpenChainGraph compute node (compliance_mandate).2 findings
Why it is in this group
- The name "validate_c2pa_manifest" says "validate": a word for reading.
Findings
- Info
  The title mentions "Credential".
  Matched text
  title: "C2PA Content Credential Manifest Validator"
  What this check looks for
- Info
  The description mentions "Credential".
  Matched text
  description: "C2PA Content Credential Manifest Validator: OpenChainGraph comp..."
  What this check looks for
Description, as the server wrote it
C2PA Content Credential Manifest Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-124-content-credential-signature-verifier. Open at: https://ainumbers.co/chaingraph/art-123-c2pa-manifest-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_canton_dvp_atomicityCanton DvP Atomicity Validator: OpenChainGraph compute node (settlement_mandate).No findings
Why it is in this group
- The name "validate_canton_dvp_atomicity" says "validate": a word for reading.
Description, as the server wrote it
Canton DvP Atomicity Validator: OpenChainGraph compute node (settlement_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: 505-tokenized-collateral-eligibility-checker. Open at: https://ainumbers.co/tools/507-canton-dvp-atomicity-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate atomic DvP settlement on Canton — PFMI P12, finality model, unwind semantics. Outputs: verdict, atomicity_flag, finality_flag, herstatt_flag, compliance_flags, pacs008, execution_hash. Output schema: call describe_tool("validate_canton_dvp_atomicity").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_canton_party_allowlistCanton Party Allowlist Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_canton_party_allowlist" says "validate": a word for reading.
Description, as the server wrote it
Canton Party Allowlist Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/tools/509-canton-party-allowlist-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Screen parties against FATF Travel Rule requirements and AML/KYA checks for Canton Network onboarding. Outputs: party_results, portfolio_verdict, compliance_flags, iso20022_party_identification. Output schema: call describe_tool("validate_canton_party_allowlist").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_canton_selective_disclosureCanton Selective-Disclosure DvP Reconciliation Attestation: OpenChainGraph compute node (attestation_mandate).No findings
Why it is in this group
- The name "validate_canton_selective_disclosure" says "validate": a word for reading.
Description, as the server wrote it
Canton Selective-Disclosure DvP Reconciliation Attestation: OpenChainGraph compute node (attestation_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: 507-canton-dvp-atomicity-validator. Output feeds: cry-01-zk-compliance-proof-generator. Open at: https://ainumbers.co/chaingraph/art-108-canton-selective-disclosure.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Attest that a Canton DvP privacy partition is sound: each counterparty sees only its leg, no cross-leg data leaks, and both views reconcile to one shared commitment. Outputs: bank_view_ok, cross_leak_fields, no_cross_leg_leak, partition_attestation, reconciles_to_commitment, registrar_view_ok, verdict. Output schema: call describe_tool("validate_canton_selective_disclosure").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_cat_bond_trigger_termsCat Bond Trigger Terms Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_cat_bond_trigger_terms" says "validate": a word for reading.
Description, as the server wrote it
Cat Bond Trigger Terms Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-251-compute-parametric-trigger-payout. Open at: https://ainumbers.co/chaingraph/art-252-validate-cat-bond-trigger-terms.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates catastrophe bond trigger term structure and computes layer arithmetic: attachment/exhaustion point ordering, pro-rata layer penetration factor, payout amount, and layer position (BELOW_ATTACHMENT / WITHIN_LAYER / ABOVE_EXHAUSTION). Outputs: attachment_breached, cascade_attachment_check, coverage_amount_used, excess_above_attachment, exhaustion_reached, implied_coverage, issues, layer_position, layer_width, not_legal_advice, payout_amount, pii_note, pro_rata_factor, regulatory_basis, table_source, table_version, terms_valid. Output schema: call describe_tool("validate_cat_bond_trigger_terms").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_cctp_v2_transferArc CCTP v2 Transfer Validator: OpenChainGraph compute node (settlement_mandate).No findings
Why it is in this group
- The name "validate_cctp_v2_transfer" says "validate": a word for reading.
Description, as the server wrote it
Arc CCTP v2 Transfer Validator: OpenChainGraph compute node (settlement_mandate). Regulatory deadline: 2026-10-31 (CCTP V1 (Legacy) deprecation begins 31 Oct 2026 and completes 1 Dec 2026 (Circle migration guide, read 1 Oct 2026). All V1 integrations must migrate to CCTP v2. The v1_sunset check output still cites the earlier 31 Jul 2026 date.). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-42-arc-fit-diagnostic. Open at: https://ainumbers.co/chaingraph/art-47-arc-cctp-transfer.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates a CCTP v2 cross-chain USDC transfer for domain pair eligibility, Fast Transfer 30-second finality risk (LP availability), Hook payload safety, CCTP V1 (Legacy) deprecation status (begins 31 Oct 2026, completes 1 Dec 2026), and large-notional LP-depth risk. Outputs: verdict, grade, fail_count, warn_count, source_domain, dest_domain, transfer_mode, notional_usd, compliance_flags. Output schema: call describe_tool("validate_cctp_v2_transfer").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_collateral_swap_eligibilityCollateral Swap Eligibility Validator: OpenChainGraph compute node (collateral_mandate).No findings
Why it is in this group
- The name "validate_collateral_swap_eligibility" says "validate": a word for reading.
Description, as the server wrote it
Collateral Swap Eligibility Validator: OpenChainGraph compute node (collateral_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: 505-tokenized-collateral-eligibility-checker, 507-canton-dvp-atomicity-validator. Open at: https://ainumbers.co/tools/515-collateral-swap-eligibility-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate a collateral swap (upgrade/downgrade trade) under GMSLA, GMRA, and SFTR Art 15 collateral reuse rules. Outputs: eligibility, hqla_impact, hqla_tier_a, hqla_tier_b, net_economic_value, pacs008, value_a, value_b. Output schema: call describe_tool("validate_collateral_swap_eligibility").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_commission_hierarchyCommission Hierarchy Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_commission_hierarchy" says "validate": a word for reading.
Description, as the server wrote it
Commission Hierarchy Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-266-reconcile-commission-statement. Open at: https://ainumbers.co/chaingraph/art-264-validate-commission-hierarchy.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: BFS structural validation of multi-level sales commission hierarchies. Outputs: agent_count, by_level, is_valid, not_legal_advice, orphan_count, override_stacking_detected, pii_note, regulatory_basis, table_source, table_version, total_levels, violations. Output schema: call describe_tool("validate_commission_hierarchy").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_content_binding_assertionContent Binding Assertion Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_content_binding_assertion" says "validate": a word for reading.
Description, as the server wrote it
Content Binding Assertion Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-127-dual-layer-disclosure-verifier. Open at: https://ainumbers.co/chaingraph/art-128-content-binding-assertion-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_cross_network_settlementCross-Network Atomic Settlement Validator: OpenChainGraph compute node (settlement_mandate).No findings
Why it is in this group
- The name "validate_cross_network_settlement" says "validate": a word for reading.
Description, as the server wrote it
Cross-Network Atomic Settlement Validator: OpenChainGraph compute node (settlement_mandate). Regulatory deadline: 2026-Q3 (ECB Pontes TARGET-link pilot end-Q3 2026; DTCC Collateral AppChain full production Oct 2026. Verify cross-network coordination patterns against current primary sources.). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-56-tokenized-settlement-fit-diagnostic, art-59-settlement-asset-finality-classifier. Output feeds: 507-canton-dvp-atomicity-validator, 511-multi-currency-pvp-validator, cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-58-cross-network-settlement-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates atomic settlement across two or more networks: cash leg final on the money ledger, asset leg delivered on the asset ledger, FX leg PvP where present. Outputs: atomicity_verdict, coordination_recommendation, leg_findings, note, pvp_check, residual_exposure, settlement_risk_window_sec. Output schema: call describe_tool("validate_cross_network_settlement").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_cyclonedx_sbomCycloneDX SBOM Validator (EU CRA Annex I): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_cyclonedx_sbom" says "validate": a word for reading.
Description, as the server wrote it
CycloneDX SBOM Validator (EU CRA Annex I): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-136-slsa-provenance-verifier. Open at: https://ainumbers.co/chaingraph/art-135-cyclonedx-sbom-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates a CycloneDX SBOM against the EU CRA Annex I machine-readable SBOM requirement: bomFormat=CycloneDX, specVersion in [1.4,1.5,1.6], all components carry purl+name+version, top-level dependencies array present. Outputs: component_count, components_missing_purl, format, has_dependencies, sbom_valid, spec_version. Output schema: call describe_tool("validate_cyclonedx_sbom").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_deposit_token_complianceDeposit-Token Compliance Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_deposit_token_compliance" says "validate": a word for reading.
Description, as the server wrote it
Deposit-Token Compliance Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-56-tokenized-settlement-fit-diagnostic. Output feeds: 510-digital-asset-regulatory-classifier, cry-04-merkle-batch-verifier, art-59-settlement-asset-finality-classifier. Open at: https://ainumbers.co/chaingraph/art-57-deposit-token-compliance-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: 3-test validator distinguishing a bank-liability deposit token (JPMD/RLN model: at-par-on-demand, on-balance-sheet, allowlisted-wholesale) from a reserve-backed stablecoin or e-money token. Outputs: applicable_regime, capital_accounting_note, classification, classification_grade, note, remediation_checklist, status_asof, test_results, token_class. Output schema: call describe_tool("validate_deposit_token_compliance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_dpp_data_carrierEU ESPR Digital Product Passport Data Carrier Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_dpp_data_carrier" says "validate": a word for reading.
Description, as the server wrote it
EU ESPR Digital Product Passport Data Carrier Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-116-product-lineage-builder. Open at: https://ainumbers.co/chaingraph/art-115-dpp-data-carrier-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate DPP required data elements against the CIRPASS-2 Core Ontology (durability, reparability, recyclability, carbon footprint, substances of concern) and check GS1 Digital Link data-carrier type. Outputs: carrier_valid, missing_elements, ontology_conformant, ontology_version, product_id. Output schema: call describe_tool("validate_dpp_data_carrier").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_dtc_tokenized_treasuryDTC-Custodied Tokenized U.S.No findings
Why it is in this group
- The name "validate_dtc_tokenized_treasury" says "validate": a word for reading.
Description, as the server wrote it
DTC-Custodied Tokenized U.S. Treasury Issuance & DvP: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: 510-digital-asset-regulatory-classifier. Output feeds: 507-canton-dvp-atomicity-validator. Open at: https://ainumbers.co/chaingraph/art-109-dtc-tokenized-treasury.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate a DTCC/ComposerX tokenized U.S. Outputs: collateral_reuse_ok, custody_link_ok, daml_lifecycle_gaps, daml_template_ok, dvp_ready, fed_eligible, verdict. Output schema: call describe_tool("validate_dtc_tokenized_treasury").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_ebam_acmt_floweBAM Account Message Flow Validation: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_ebam_acmt_flow" says "validate": a word for reading.
Description, as the server wrote it
eBAM Account Message Flow Validation: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-260-allocate-ihb-interest. Open at: https://ainumbers.co/chaingraph/art-262-validate-ebam-acmt-flow.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates the CGI-MP eBAM 2023 account message state machine across acmt.007 (opening request), acmt.010 (opening confirmation), acmt.011 (closing request), acmt.017 (modification request), and acmt.019 (modification confirmation) message types. Outputs: ack_count, acmt_state, error_count, is_valid, message_sequence, not_legal_advice, orphan_count, orphan_requests, pii_note, regulatory_basis, report_count, request_count, table_source, table_version, total_messages, validation_errors, validation_warnings, warning_count. Output schema: call describe_tool("validate_ebam_acmt_flow").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_einvoice_batchEN 16931 / Factur-X E-Invoicing Batch Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_einvoice_batch" says "validate": a word for reading.
Description, as the server wrote it
EN 16931 / Factur-X E-Invoicing Batch Validator: OpenChainGraph compute node (compliance_mandate). Regulatory deadline: 2026-09 (France large/medium mandatory September 2026; EU ViDA). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: rca-03-iso20022-address-migration-verifier, ptg-01-ap2-prompt-template-generator. Open at: https://ainumbers.co/chaingraph/art-08-en16931-einvoice-batch-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_einvoice_formatE-Invoice Format Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_einvoice_format" says "validate": a word for reading.
Description, as the server wrote it
E-Invoice Format Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-294-einvoice-vat-calc-verifier. Open at: https://ainumbers.co/chaingraph/art-293-einvoice-format-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Conformance-validate a structured e-invoice extract against version-pinned Factur-X, XRechnung, PINT-AE, MyInvois, Peppol BIS 3.0/EN 16931 core (Belgium), or KSeF FA(3) (Poland) format rules: mandatory-field presence, currency/VAT-category codelist membership, line-item cardinality. Outputs: findings, format, line_item_count, missing_fields, parse_error, rule_set_version, structural_completeness. Output schema: call describe_tool("validate_einvoice_format").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_emir_lifecycle_eventEMIR Lifecycle Event Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_emir_lifecycle_event" says "validate": a word for reading.
Description, as the server wrote it
EMIR Lifecycle Event Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-156-emir-counterparty-pairing-reconciler. Output feeds: art-158-emir-reporting-readiness-diagnostic. Open at: https://ainumbers.co/chaingraph/art-157-emir-lifecycle-event-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_emir_trade_reportEMIR Trade Report Field Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_emir_trade_report" says "validate": a word for reading.
Description, as the server wrote it
EMIR Trade Report Field Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-154-emir-uti-completeness-checker. Open at: https://ainumbers.co/chaingraph/art-153-emir-trade-report-field-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_emir_upiEMIR UPI Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_emir_upi" says "validate": a word for reading.
Description, as the server wrote it
EMIR UPI Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-154-emir-uti-completeness-checker. Open at: https://ainumbers.co/chaingraph/art-155-emir-upi-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_eudr_due_diligence_statementEUDR DDS Field Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_eudr_due_diligence_statement" says "validate": a word for reading.
Description, as the server wrote it
EUDR DDS Field Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-166-eudr-geolocation-plot-validator. Open at: https://ainumbers.co/chaingraph/art-165-eudr-dds-field-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate the required-field subset of an EUDR Due Diligence Statement (DDS) before TRACES NT filing: operator name, address, EORI, HS code, trade name, quantity, country of production, and geolocation indicator (or micro-operator postal-address exemption). Outputs: conformant, country_of_production, fields_checked, fields_passed, micro_operator_exemption, missing_fields, quantity. Output schema: call describe_tool("validate_eudr_due_diligence_statement").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_eudr_geolocationEUDR Geolocation Plot Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_eudr_geolocation" says "validate": a word for reading.
Description, as the server wrote it
EUDR Geolocation Plot Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-165-eudr-dds-field-validator. Output feeds: art-167-eudr-commodity-scope-classifier. Open at: https://ainumbers.co/chaingraph/art-166-eudr-geolocation-plot-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate GeoJSON plot geolocation for EUDR compliance: geometry type (Point or Polygon), coordinate range validity, EUDR size rule (plots >=4 ha require Polygon, Art. Outputs: area_ha, coordinates_valid, geo_type, issues, micro_operator_exemption, polygon_closed, size_rule_met, valid. Output schema: call describe_tool("validate_eudr_geolocation").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_eugb_factsheetEU Green Bond Factsheet & Allocation Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_eugb_factsheet" says "validate": a word for reading.
Description, as the server wrote it
EU Green Bond Factsheet & Allocation Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-68-carbon-compliance-fit-diagnostic, art-73-taxonomy-alignment-scorer. Output feeds: cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-75-eugb-factsheet-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates an EuGB factsheet (Annex I) + allocation report (Annex II) for completeness and the 100% Taxonomy-aligned proceeds threshold, cross-checking against ART-73 alignment of the funded activities. Outputs: aligned_proceeds, annex_i_complete, annex_i_gaps, annex_ii_status, conformance_grade, conformance_score, external_reviewer_status, label_ready, note, proceeds_aligned_pct, proceeds_threshold_met, proceeds_threshold_pct, reference, total_proceeds. Output schema: call describe_tool("validate_eugb_factsheet").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_form5500_schedulesERISA Form 5500 Schedule Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_form5500_schedules" says "validate": a word for reading.
Description, as the server wrote it
ERISA Form 5500 Schedule Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-401-validate-form5500-schedules.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates a Form 5500 schedule-applicability matrix (plan type and size determine required schedules H/I/A/C/G/MB/SB/R), a Schedule H cross-schedule arithmetic tie (ending assets equal beginning assets plus net income minus distributions), and the filing-deadline calculation (plan-year end plus s... Outputs: applicable_deadline, arithmetic_tie, compliant, disambiguation, error_count, extended_filing_deadline, filing_deadline, is_large_plan, issues, regulatory_basis, required_schedules, shelf_note, table_source, table_version, warning_count. Output schema: call describe_tool("validate_form5500_schedules").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_fsma204_cteFSMA 204 Critical Tracking Event (CTE) Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_fsma204_cte" says "validate": a word for reading.
Description, as the server wrote it
FSMA 204 Critical Tracking Event (CTE) Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-119-traceability-lot-code-linker. Open at: https://ainumbers.co/chaingraph/art-118-fsma204-cte-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate required Key Data Elements present for each FDA FSMA 204 Critical Tracking Event (harvesting/cooling/initial packing/shipping/receiving/transformation) per the Food Traceability List. Outputs: cte_type, cte_valid, ftl_food, missing_kdes. Output schema: call describe_tool("validate_fsma204_cte").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_fund_collateralTokenized Fund Collateral Validator: OpenChainGraph compute node (collateral_mandate).No findings
Why it is in this group
- The name "validate_fund_collateral" says "validate": a word for reading.
Description, as the server wrote it
Tokenized Fund Collateral Validator: OpenChainGraph compute node (collateral_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: 505-tokenized-collateral-eligibility-checker. Open at: https://ainumbers.co/tools/514-tokenized-fund-collateral-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate tokenized MMF / CNAV / LVNAV / VNAV fund shares as collateral. Outputs: adjusted_collateral_value, eligibility, haircut_applied, hqla_tier. Output schema: call describe_tool("validate_fund_collateral").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_ifrs17_csm_rollforwardIFRS 17 CSM Roll-Forward Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_ifrs17_csm_rollforward" says "validate": a word for reading.
Description, as the server wrote it
IFRS 17 CSM Roll-Forward Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-177-ifrs17-measurement-model-classifier. Output feeds: art-179-ifrs17-risk-adjustment-checker. Open at: https://ainumbers.co/chaingraph/art-178-ifrs17-csm-rollforward-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate IFRS 17 Contractual Service Margin (CSM) roll-forward mechanics: opening CSM + new business + interest accretion + experience adjustments - coverage-unit release + FX adjustments = closing CSM. Outputs: closing_csm, csm_valid, experience_adjustments, fx_adjustments, interest_accretion, loss_component, new_business_csm, onerous, opening_csm, release_to_profit. Output schema: call describe_tool("validate_ifrs17_csm_rollforward").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_input_attestationsVerify an artifact's input_attestations[] (ChainGraph Standard §23): per RFC 6901 pointer, checks the attested value resolves inside policy_parameters and its digest binding matche…No findings
Why it is in this group
- The name "validate_input_attestations" says "validate": a word for reading.
Description, as the server wrote it
Verify an artifact's input_attestations[] (ChainGraph Standard §23): per RFC 6901 pointer, checks the attested value resolves inside policy_parameters and its digest binding matches, then verifies each type along its own path -- vc-2.0 via the shipped §16/§13.11 Data Integrity proof, rfc3161-snapshot via the same §20 rfc3161-tst verifier (no second RFC 3161 implementation), c2pa-manifest structurally (hard-binding digest match), zktls structurally-only (reported verifiable:"external" -- OCG never treats it as confirmed). Returns one { pointer, type, structural, verifiable } record per entry. Pure client-safe compute, zero network.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_mcp_authorization_metadataMCP Authorization Metadata Validator (RFC 9728): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_mcp_authorization_metadata" says "validate": a word for reading.
Description, as the server wrote it
MCP Authorization Metadata Validator (RFC 9728): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-147-mcp-server-identity-attestation-validator. Output feeds: art-149-mcp-registry-entry-conformance. Open at: https://ainumbers.co/chaingraph/art-148-mcp-authorization-metadata-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate OAuth 2.0 Protected Resource Metadata per RFC 9728: resource URI (https-scheme), non-empty authorization_servers, scopes_supported, and bearer_methods_supported restricted to header/body/query. Outputs: auth_server_count, bearer_ok, metadata_valid, missing, resource_ok, scope_count. Output schema: call describe_tool("validate_mcp_authorization_metadata").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_mcp_server_identityMCP Server Identity Attestation Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_mcp_server_identity" says "validate": a word for reading.
Description, as the server wrote it
MCP Server Identity Attestation Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-148-mcp-authorization-metadata-validator. Open at: https://ainumbers.co/chaingraph/art-147-mcp-server-identity-attestation-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate a new-spec MCP server identity document: required claims (subject, issuer, serverInfo), well-known path correctness (/.well-known/mcp-server-identity), and attestation reference present. Outputs: attested, has_issuer, has_server_info, has_subject, identity_valid, missing. Output schema: call describe_tool("validate_mcp_server_identity").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_mcp_server_jsonValidate an MCP server.json against the 2025-12-11 schema and the official registry publishing rules; returns findings, a registry-readiness score, and an optional compliant skelet…No findings
Why it is in this group
- The name "validate_mcp_server_json" says "validate": a word for reading.
Description, as the server wrote it
Validate an MCP server.json against the 2025-12-11 schema and the official registry publishing rules; returns findings, a registry-readiness score, and an optional compliant skeleton. Use when a developer wants to check a server.json before publishing to the MCP Registry. Renders the interactive AINumbers tool as a widget; inputs are applied via the AIN Bridge and the tool runs client-side (zero PII, zero network). Output schema: call describe_tool("validate_mcp_server_json").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_mcp_task_lifecycleMCP Task Lifecycle State Machine Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_mcp_task_lifecycle" says "validate": a word for reading.
Description, as the server wrote it
MCP Task Lifecycle State Machine Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-151-agent-obo-mandate-validator. Open at: https://ainumbers.co/chaingraph/art-152-mcp-task-lifecycle-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate that a long-running MCP task state transitions are legal per the new MCP specification state machine: working to input_required or terminal states (completed, failed, cancelled); input_required back to working or terminal. Outputs: illegal_transitions, lifecycle_valid, transition_count. Output schema: call describe_tool("validate_mcp_task_lifecycle").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_mletr_recordMLETR / eBL Conformance & Enforceability Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_mletr_record" says "validate": a word for reading.
Description, as the server wrote it
MLETR / eBL Conformance & Enforceability Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-52-digital-trade-fit-diagnostic. Output feeds: 510-digital-asset-regulatory-classifier, cry-04-merkle-batch-verifier, ml-02-credit-default-risk-scorer. Open at: https://ainumbers.co/chaingraph/art-53-mletr-ebl-conformance-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates an electronic transferable record (eBL or other ETR) against MLETR functional-equivalence tests (Arts. Outputs: conformance_grade, conformance_score, corridor_matrix, enforceability_tier, governing_law_recommendation, note, remediation_checklist, status_table_asof, test_results. Output schema: call describe_tool("validate_mletr_record").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_nft_metadata_art209NFT Metadata Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_nft_metadata_art209" says "validate": a word for reading.
Description, as the server wrote it
NFT Metadata Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-209-nft-metadata-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates NFT metadata against ERC-721 and ERC-1155 required fields plus OpenSea metadata schema recommendations. Outputs: all_pass, checks, disclaimer, fail_count, field_count, required_pass, valid, warn_count. Output schema: call describe_tool("validate_nft_metadata_art209").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_pacs008_party_completenesspacs.008 Party Completeness Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_pacs008_party_completeness" says "validate": a word for reading.
Description, as the server wrote it
pacs.008 Party Completeness Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-241-cbpr-structured-address-linter. Output feeds: art-246-lei-payment-binding-linter. Open at: https://ainumbers.co/chaingraph/art-242-pacs008-party-completeness-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates BIS CPMI d218 harmonised data requirements for a pacs.008 payment instruction: UETR UUIDv4 format, debtor and creditor names, BIC format, LEI format (presence and format; full mod-97 check-digit validation is in lint_lei_payment_binding art-246), and purpose code format. Outputs: compliant, cpmi_d218_score, disambiguation, error_count, field_status, issues, pii_note, regulatory_basis, table_source, table_version, warning_count. Output schema: call describe_tool("validate_pacs008_party_completeness").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_private_inputsVerify an artifact's private_inputs[] (ChainGraph Standard §25 ocg-private-input@1) WITHOUT ever seeing the plaintext witness: per RFC 6901 pointer, checks the pointed value inside…No findings
Why it is in this group
- The name "validate_private_inputs" says "validate": a word for reading.
Description, as the server wrote it
Verify an artifact's private_inputs[] (ChainGraph Standard §25 ocg-private-input@1) WITHOUT ever seeing the plaintext witness: per RFC 6901 pointer, checks the pointed value inside policy_parameters IS the declared sha256-salted@1 commitment (never the plaintext, §25.2), that the commitment scheme is known, and -- when a §18 compute_proof is present -- that its journal commits the same commitment and binds output_payload. Optionally accepts an out-of-band {pointer, salt, input_value} disclosure package (authorized-verifier path) and recomputes sha256(salt || cgCanon(input_value)) to confirm it equals the commitment. Returns one {pointer, verifiable} record per entry: "proof-only" | "disclosed-verified" | "commitment-only" (no proof yet -- structural + plaintext-exclusion only) | "failed". Pure client-safe compute, zero network, never requires the plaintext.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_pvp_settlementMulti-Currency PvP Validator: OpenChainGraph compute node (settlement_mandate).No findings
Why it is in this group
- The name "validate_pvp_settlement" says "validate": a word for reading.
Description, as the server wrote it
Multi-Currency PvP Validator: OpenChainGraph compute node (settlement_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: 507-canton-dvp-atomicity-validator, 505-tokenized-collateral-eligibility-checker. Open at: https://ainumbers.co/tools/511-multi-currency-pvp-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate atomic Payment-vs-Payment (PvP) settlement across currency pairs. Outputs: has_canton_leg. Output schema: call describe_tool("validate_pvp_settlement").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_regf_call_frequencyReg F Call-Frequency Presumption Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_regf_call_frequency" says "validate": a word for reading.
Description, as the server wrote it
Reg F Call-Frequency Presumption Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-403-check-debt-validation-notice. Open at: https://ainumbers.co/chaingraph/art-402-validate-regf-call-frequency.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Checks a declared debt-collection call log against the two 12 CFR 1006.14(b) Regulation F rebuttable presumptions: more than seven telephone calls to a person on a single debt within any seven consecutive days, and a call placed within seven days after a telephone conversation with that person on... Outputs: debts, debts_checked, debts_with_quiet_period_presumption, debts_with_seven_in_seven_presumption, disambiguation, invalid_call_indices, regulatory_basis, timezone_offset_minutes_applied, window_days. Output schema: call describe_tool("validate_regf_call_frequency").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_royalty_splitRoyalty Split Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_royalty_split" says "validate": a word for reading.
Description, as the server wrote it
Royalty Split Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-208-royalty-split-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates on-chain royalty split configurations against ERC-2981 (single-receiver) and 0xSplits (multi-recipient) rules. Outputs: cap_bps, config_hash, disclaimer, mode, recipient_count, rules, sum, valid. Output schema: call describe_tool("validate_royalty_split").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_signature_agent_cardSignature Agent Card Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_signature_agent_card" says "validate": a word for reading.
Description, as the server wrote it
Signature Agent Card Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-130-signature-directory-validator. Open at: https://ainumbers.co/chaingraph/art-131-signature-agent-card-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_signature_directoryHTTP Signatures Directory Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_signature_directory" says "validate": a word for reading.
Description, as the server wrote it
HTTP Signatures Directory Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-129-webbotauth-signature-verifier. Output feeds: art-131-signature-agent-card-validator. Open at: https://ainumbers.co/chaingraph/art-130-signature-directory-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_spdx_sbomSPDX SBOM Validator (EU CRA Annex I): OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_spdx_sbom" says "validate": a word for reading.
Description, as the server wrote it
SPDX SBOM Validator (EU CRA Annex I): OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-139-cra-annex1-completeness-checker. Open at: https://ainumbers.co/chaingraph/art-138-spdx-sbom-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates an SPDX SBOM against the EU CRA Annex I machine-readable SBOM requirement: spdxVersion matches SPDX-2.x or SPDX-3.x, SPDXID present, all packages carry name+versionInfo+downloadLocation (or purl externalRef), relationships array non-empty. Outputs: format, has_relationships, package_count, packages_missing_version, sbom_valid, spdx_version. Output schema: call describe_tool("validate_spdx_sbom").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_tempo_token_complianceTempo Stablecoin Issuance Compliance: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_tempo_token_compliance" says "validate": a word for reading.
Description, as the server wrote it
Tempo Stablecoin Issuance Compliance: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-34-tempo-fit-diagnostic. Output feeds: art-06-genius-act-reserve-attestation, art-10-amla-transaction-typology-risk-scorer, art-38-tempo-onchain-aml. Open at: https://ainumbers.co/chaingraph/art-37-tempo-stablecoin-issuance.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Dual-jurisdiction TIP-20 token compliance validator. Outputs: fail_count, genius, mica, verdict, warn_count. Output schema: call describe_tool("validate_tempo_token_compliance").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_tempo_zone_disclosureTempo Zone Selective-Disclosure Attestation: OpenChainGraph compute node (attestation_mandate).No findings
Why it is in this group
- The name "validate_tempo_zone_disclosure" says "validate": a word for reading.
Description, as the server wrote it
Tempo Zone Selective-Disclosure Attestation: OpenChainGraph compute node (attestation_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-38-tempo-onchain-aml. Output feeds: cry-01-zk-compliance-proof-generator. Open at: https://ainumbers.co/chaingraph/art-39-tempo-zone-disclosure.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Maps a Tempo Zone's party-visibility model (operator-sees-all / users-see-own / outsiders-see-ZK-proofs) against AML/audit/regulator disclosure obligations. Outputs: checks, operator_name, use_case, verdict. Output schema: call describe_tool("validate_tempo_zone_disclosure").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_tfr_travel_rule_batchTFR Travel-Rule Batch Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_tfr_travel_rule_batch" says "validate": a word for reading.
Description, as the server wrote it
TFR Travel-Rule Batch Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-98-mica-casp-fit-diagnostic. Output feeds: cry-04-merkle-batch-verifier. Open at: https://ainumbers.co/chaingraph/art-104-tfr-travel-rule-batch-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates originator/beneficiary field completeness on synthetic/hashed transfer batches (self-/cross-CASP + unhosted-wallet branches) per TFR recast Reg. Outputs: batch_conformance_pct, batch_size, merkle_root, note, reference_version, tfr_note, transfers_flagged, unhosted_dd_required_count. Output schema: call describe_tool("validate_tfr_travel_rule_batch").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_tip20_memo_commitmentTIP-20 Memo/Commitment Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_tip20_memo_commitment" says "validate": a word for reading.
Description, as the server wrote it
TIP-20 Memo/Commitment Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-390-tip20-memo-commitment-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates a TIP-20 TransferWithMemo's 32-byte memo as a hash-or-locator commitment: checks length/hex form, recomputes the SHA-256 commitment over a supplied off-chain payload or a templated invoice-ID locator, and reports whether the memo matches. Outputs: commitment_source_supplied, invoice_locator, invoice_locator_commitment, invoice_locator_match, memo_hex, memo_hex_valid, memo_length_valid, note, overall_valid, payload_commitment, payload_commitment_match. Output schema: call describe_tool("validate_tip20_memo_commitment").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_tokenized_security_lifecycleTokenized Security Lifecycle Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_tokenized_security_lifecycle" says "validate": a word for reading.
Description, as the server wrote it
Tokenized Security Lifecycle Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: 510-digital-asset-regulatory-classifier. Open at: https://ainumbers.co/tools/512-tokenized-security-lifecycle-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validate issuance, corporate actions, and maturity/redemption lifecycle for tokenized securities on Daml/Canton. Outputs: all_gaps, critical_gaps, event_matrix, verdict, verdict_badge. Output schema: call describe_tool("validate_tokenized_security_lifecycle").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_vida_einvoice_conformanceViDA EN 16931 E-Invoice Conformance Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_vida_einvoice_conformance" says "validate": a word for reading.
Description, as the server wrote it
ViDA EN 16931 E-Invoice Conformance Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Output feeds: art-160-vida-drr-transaction-reporter. Open at: https://ainumbers.co/chaingraph/art-159-vida-einvoice-en16931-conformance-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_w8_series_structuralW-8 Series Structural Validator: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "validate_w8_series_structural" says "validate": a word for reading.
Description, as the server wrote it
W-8 Series Structural Validator: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-268-compute-cdd-ownership-25pct. Open at: https://ainumbers.co/chaingraph/art-269-validate-w8-series-structural.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates W-8 series form structural consistency for withholding tax compliance. Outputs: ch3_ch4_consistent, chapter3_status, chapter4_fatca_status, days_until_expiry, form_ch3_compatible, form_type, is_structurally_valid, not_legal_advice, pii_note, regulatory_basis, statutory_withholding_rate_pct, table_source, table_version, treaty_country, treaty_rate_expected, treaty_rate_pct, treaty_rate_valid, validity_expiry_date, validity_window_ok, violation_count, violations. Output schema: call describe_tool("validate_w8_series_structural").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
validate_x402_deferred_handshakex402 Deferred-Scheme Handshake Validator: OpenChainGraph compute node (compliance_control).No findings
Why it is in this group
- The name "validate_x402_deferred_handshake" says "validate": a word for reading.
Description, as the server wrote it
x402 Deferred-Scheme Handshake Validator: OpenChainGraph compute node (compliance_control). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-393-x402-v2-migration-linter, art-129-webbotauth-signature-verifier. Output feeds: art-61-x402-batch-settlement-reconciler. Open at: https://ainumbers.co/chaingraph/art-394-x402-deferred-handshake-validator.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Validates a Cloudflare `deferred` x402 scheme handshake: 402 offer field shape (scheme:"deferred", id, termsUrl), RFC 9421 HTTP Message Signature covered-component coverage (@method, @target-uri, content-digest), and settlement-reference id continuity across a call sequence (uniqueness, no intern... Outputs: errors, findings, passes, required_covered_components, scope_note, score, settlement_reference_id, verdict, warnings. Output schema: call describe_tool("validate_x402_deferred_handshake").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_did_webvh_logdid:webvh DID Log Verifier: OpenChainGraph compute node (compliance_mandate).No findings
Why it is in this group
- The name "verify_did_webvh_log" says "log": a word for reading.
Description, as the server wrote it
did:webvh DID Log Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Consumes upstream artifacts from: art-04-agent-identity-attestation-checker. Output feeds: art-285-acdc-delegation-chain-verifier. Open at: https://ainumbers.co/chaingraph/art-284-did-webvh-log-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Verifies a did:webvh self-certifying DID log: per-entry self-hash integrity, sequential versionId numbering, update-key-authorized Ed25519 signatures on every entry, deactivation status, and optional resolved-document match. Outputs: current_version_id, deactivated, did, entries_checked, failures, valid. Output schema: call describe_tool("verify_did_webvh_log").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
verify_revocation_statusRevocation-Status Verifier: OpenChainGraph compute node (compliance_mandate).1 finding
Why it is in this group
- The name "verify_revocation_status" says "status": a word for reading.
Findings
- Info
  The description mentions "credential".
  Matched text
  description: "...rom a supplied, zero-egress status list credential. Outputs: note, revoked_for_purpose, st..."
  What this check looks for
Description, as the server wrote it
Revocation-Status Verifier: OpenChainGraph compute node (compliance_mandate). Deterministic OpenChainGraph compute node. By default (compute:"auto") inputs are computed server-side on Cloudflare Workers for gpu:false nodes with a registered kernel; compute:"browser" forces client-side execution and returns a browser delegation URL instead. gpu:true nodes always delegate to the browser. Inputs are processed transiently to compute the response and are not stored, logged, or retained. Use synthetic or anonymised inputs only. Exports an AP2 artifact with execution_hash for chain provenance. Open at: https://ainumbers.co/chaingraph/art-287-revocation-status-verifier.html FV-status (published/proven/still-trusted for this spec): /fv-status/27ceec007c8035426d8ad4b2ade603d763c5b223d5a2e86369e7c27af06fca2b.json — a snapshot, not a subscription; this receipt verifies offline regardless of whether that file is ever fetched. Summary: Checks a receipt's optional W3C BitstringStatusList credentialStatus reference and reads the revocation bit at statusListIndex from a supplied, zero-egress status list credential. Outputs: note, revoked_for_purpose, status, status_list_credential_url, status_list_index, structural_error. Output schema: call describe_tool("verify_revocation_status").
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
workbook_roundtrip_verifyCompares a WB-2 Spreadsheet Input Manifest (expected digests) against pasted-back CSV/TSV text per manifest range (observed, e.g.No findings
Why it is in this group
- The description's first verb is "compare": a word for reading.
Description, as the server wrote it
Compares a WB-2 Spreadsheet Input Manifest (expected digests) against pasted-back CSV/TSV text per manifest range (observed, e.g. after a recompute in Excel) and returns an XLR-1 round-trip receipt -- `result: "match"|"mismatch"` plus a `mismatches[]` cell list when expected_by_ref text is also supplied. SAME comparator module as the tools/ round-trip page (XLR-2/XLR-3) -- byte-identical receipt for the same inputs. Paste-intake is untrusted: finite-gate (#NUM! for NaN/Infinity) and CSV-injection sanitization apply identically to WB-1's CSV import. Verify-only -- never operates Excel, never ingests .xlsx.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false

This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.

Install

Pinned to version 0.4.26. Values in angle brackets are yours to fill in.

claude mcp add --transport http tools https://mcp.ainumbers.co/mcp

Run in a terminal.

Claude Code permissions

Allow the reads, ask before a change, deny the deletes. Paste into .claude/settings.json.

.claude/settings.json

{
  "permissions": {
    "allow": [
      "mcp__tools__list_ainumbers_tools",
      "mcp__tools__find_tool",
      "mcp__tools__find_chain",
      "mcp__tools__describe_tool",
      "mcp__tools__validate_ap2_mcp_policy",
      "mcp__tools__lint_mcp_tool_definition",
      "mcp__tools__validate_mcp_server_json",
      "mcp__tools__compare_agentic_payment_protocols",
      "mcp__tools__validate_a2a_agent_card",
      "mcp__tools__inspect_visa_tap_signature",
      "mcp__tools__validate_input_attestations",
      "mcp__tools__validate_private_inputs",
      "mcp__tools__checklist_validate_definition",
      "mcp__tools__ha_record_validate",
      "mcp__tools__ha_gate_status",
      "mcp__tools__otlp_validate",
      "mcp__tools__redline_diff",
      "mcp__tools__lei_kyb_check",
      "mcp__tools__workbook_roundtrip_verify",
      "mcp__tools__pain001_validate",
      "mcp__tools__validate_canton_dvp_atomicity",
      "mcp__tools__validate_canton_party_allowlist",
      "mcp__tools__validate_pvp_settlement",
      "mcp__tools__validate_tokenized_security_lifecycle",
      "mcp__tools__validate_fund_collateral",
      "mcp__tools__validate_collateral_swap_eligibility",
      "mcp__tools__validate_ap2_mandate_chain",
      "mcp__tools__validate_einvoice_batch",
      "mcp__tools__validate_acp_checkout",
      "mcp__tools__validate_ap2_mandate_credential",
      "mcp__tools__compare_agentic_rail_protocols",
      "mcp__tools__inspect_visa_trusted_agent_protocol",
      "mcp__tools__validate_agent_commerce_conformance",
      "mcp__tools__validate_a2a_x402_mandate",
      "mcp__tools__validate_a2a_trust_chain",
      "mcp__tools__validate_tempo_token_compliance",
      "mcp__tools__validate_tempo_zone_disclosure",
      "mcp__tools__validate_cctp_v2_transfer",
      "mcp__tools__validate_mletr_record",
      "mcp__tools__validate_deposit_token_compliance",
      "mcp__tools__validate_cross_network_settlement",
      "mcp__tools__resolve_cbam_default_value",
      "mcp__tools__validate_eugb_factsheet",
      "mcp__tools__check_screening_list_coverage",
      "mcp__tools__validate_tfr_travel_rule_batch",
      "mcp__tools__validate_canton_selective_disclosure",
      "mcp__tools__validate_dtc_tokenized_treasury",
      "mcp__tools__assess_suspect_product_status",
      "mcp__tools__validate_dpp_data_carrier",
      "mcp__tools__validate_fsma204_cte",
      "mcp__tools__resolve_recall_trace",
      "mcp__tools__validate_c2pa_manifest",
      "mcp__tools__resolve_provenance_ingredient_tree",
      "mcp__tools__validate_content_binding_assertion",
      "mcp__tools__validate_signature_directory",
      "mcp__tools__validate_signature_agent_card",
      "mcp__tools__validate_cyclonedx_sbom",
      "mcp__tools__validate_spdx_sbom",
      "mcp__tools__validate_mcp_server_identity",
      "mcp__tools__validate_mcp_authorization_metadata",
      "mcp__tools__validate_agent_obo_mandate",
      "mcp__tools__validate_mcp_task_lifecycle",
      "mcp__tools__validate_emir_trade_report",
      "mcp__tools__validate_emir_upi",
      "mcp__tools__validate_emir_lifecycle_event",
      "mcp__tools__validate_vida_einvoice_conformance",
      "mcp__tools__validate_eudr_due_diligence_statement",
      "mcp__tools__validate_eudr_geolocation",
      "mcp__tools__validate_ai_impact_assessment",
      "mcp__tools__validate_ifrs17_csm_rollforward",
      "mcp__tools__compare_rights_matrix",
      "mcp__tools__validate_royalty_split",
      "mcp__tools__validate_nft_metadata_art209",
      "mcp__tools__analyze_prediction_market",
      "mcp__tools__find_prediction_arbitrage",
      "mcp__tools__lookup_reg_z_thresholds",
      "mcp__tools__validate_adverse_action_notice",
      "mcp__tools__build_ai_decision_log_record",
      "mcp__tools__validate_agent_audit_trail",
      "mcp__tools__validate_pacs008_party_completeness",
      "mcp__tools__compare_corridor_cost",
      "mcp__tools__validate_cat_bond_trigger_terms",
      "mcp__tools__validate_ebam_acmt_flow",
      "mcp__tools__validate_commission_hierarchy",
      "mcp__tools__validate_w8_series_structural",
      "mcp__tools__compare_pension_lump_sum_annuity",
      "mcp__tools__verify_did_webvh_log",
      "mcp__tools__verify_revocation_status",
      "mcp__tools__validate_einvoice_format",
      "mcp__tools__lookup_mletr_jurisdiction_adoption",
      "mcp__tools__compare_basel_2023_vs_2026",
      "mcp__tools__validate_tip20_memo_commitment",
      "mcp__tools__validate_x402_deferred_handshake",
      "mcp__tools__validate_form5500_schedules",
      "mcp__tools__validate_regf_call_frequency",
      "mcp__tools__compare_model_outcome_analysis"
    ],
    "ask": [
      "mcp__tools__build_workflow_links",
      "mcp__tools__verify_execution_hash",
      "mcp__tools__build_chaingraph",
      "mcp__tools__emit_chaingraph_artifact",
      "mcp__tools__build_session_receipt",
      "mcp__tools__run_chain",
      "mcp__tools__suite_howto",
      "mcp__tools__run_chain_batch",
      "mcp__tools__call_tool",
      "mcp__tools__baas_provider_comparator",
      "mcp__tools__build_google_ap2_mandate",
      "mcp__tools__score_mcp_readiness",
      "mcp__tools__agentic_mandate_sandbox",
      "mcp__tools__customer_risk_rating",
      "mcp__tools__ap2_aml_mandate_builder",
      "mcp__tools__decode_x402_payment",
      "mcp__tools__audit_mcp_oauth",
      "mcp__tools__scan_tool_poisoning",
      "mcp__tools__run_kernel_vm",
      "mcp__tools__build_disclosure_manifest",
      "mcp__tools__verify_disclosure_inclusion",
      "mcp__tools__suggest_tool_idea",
      "mcp__tools__export_artifact",
      "mcp__tools__vc_issue",
      "mcp__tools__sdjwt_issue",
      "mcp__tools__sdjwt_present",
      "mcp__tools__checklist_step_receipt",
      "mcp__tools__checklist_verify_run",
      "mcp__tools__ha_bundle_export",
      "mcp__tools__intoto_record_chain_run",
      "mcp__tools__build_evidence_pack",
      "mcp__tools__anchor_stamp",
      "mcp__tools__otlp_span_receipt",
      "mcp__tools__redline_verify",
      "mcp__tools__acdc_said_check",
      "mcp__tools__workbook_evaluate",
      "mcp__tools__workbook_range_digest",
      "mcp__tools__workbook_csv_parse",
      "mcp__tools__camt053_parse",
      "mcp__tools__recon_match",
      "mcp__tools__diagnose_canton_readiness",
      "mcp__tools__optimize_settlement_capital",
      "mcp__tools__check_tokenized_collateral_eligibility",
      "mcp__tools__check_cash_leg_finality",
      "mcp__tools__calculate_repo_haircut",
      "mcp__tools__classify_digital_asset_regulatory",
      "mcp__tools__mobilize_margin_collateral",
      "mcp__tools__audit_borrower_escrow_stmt",
      "mcp__tools__servicing_compliance_clock",
      "mcp__tools__simulate_spend_policy",
      "mcp__tools__model_x402_settlement",
      "mcp__tools__check_agent_attestation",
      "mcp__tools__assess_ai_act_conformity",
      "mcp__tools__precheck_reserve_attestation",
      "mcp__tools__compute_basel31_delta",
      "mcp__tools__classify_dora_incident",
      "mcp__tools__score_aml_typologies",
      "mcp__tools__simulate_vop_matching",
      "mcp__tools__check_eudi_readiness",
      "mcp__tools__assess_psd3_readiness",
      "mcp__tools__simulate_agent_spend_policy",
      "mcp__tools__draft_ap2_mandate_credential",
      "mcp__tools__score_mcp_server_readiness",
      "mcp__tools__select_agentic_checkout_protocol",
      "mcp__tools__audit_acp_ucp_product_feed",
      "mcp__tools__build_agent_traffic_policy",
      "mcp__tools__build_mastercard_agentic_token",
      "mcp__tools__verify_a2a_agent_card",
      "mcp__tools__simulate_x402_flow",
      "mcp__tools__run_agentic_readiness_diagnostic",
      "mcp__tools__run_mcp_deployability_diagnostic",
      "mcp__tools__run_dora_readiness_diagnostic",
      "mcp__tools__lint_mcp_server_conformance",
      "mcp__tools__run_tempo_fit_diagnostic",
      "mcp__tools__model_tempo_payment_economics",
      "mcp__tools__decode_mpp_session",
      "mcp__tools__screen_tip20_transfer_batch",
      "mcp__tools__map_tempo_settlement",
      "mcp__tools__score_tempo_validator_readiness",
      "mcp__tools__run_arc_fit_diagnostic",
      "mcp__tools__model_arc_cpn_economics",
      "mcp__tools__model_arc_stablefx_rfq",
      "mcp__tools__lint_arc_xreserve_config",
      "mcp__tools__model_arc_paymaster_economics",
      "mcp__tools__run_treasury_clearing_fit",
      "mcp__tools__model_clearing_access_economics",
      "mcp__tools__estimate_ficc_margin_netting",
      "mcp__tools__estimate_cross_margin_benefit",
      "mcp__tools__run_digital_trade_fit",
      "mcp__tools__check_digital_trade_rules",
      "mcp__tools__verify_trade_document_set",
      "mcp__tools__run_tokenized_settlement_fit",
      "mcp__tools__classify_settlement_asset_finality",
      "mcp__tools__run_agent_economy_fit",
      "mcp__tools__reconcile_x402_batch_settlement",
      "mcp__tools__verify_ap2_payment_receipt",
      "mcp__tools__model_agent_service_metering",
      "mcp__tools__run_ai_act_highrisk_fit",
      "mcp__tools__build_ai_conformity_pack",
      "mcp__tools__build_fria_monitoring_plan",
      "mcp__tools__classify_agentic_ai_risk",
      "mcp__tools__run_carbon_compliance_fit",
      "mcp__tools__calculate_cbam_embedded_emissions",
      "mcp__tools__model_cbam_certificate_cost",
      "mcp__tools__aggregate_cbam_precursor_emissions",
      "mcp__tools__score_taxonomy_alignment",
      "mcp__tools__aggregate_taxonomy_kpi_gar",
      "mcp__tools__apply_climate_scenario",
      "mcp__tools__run_t1_readiness_diagnostic",
      "mcp__tools__calculate_csdr_penalty",
      "mcp__tools__predict_settlement_fail",
      "mcp__tools__check_ssi_conformance",
      "mcp__tools__check_allocation_affirmation",
      "mcp__tools__lint_securities_settlement_message",
      "mcp__tools__model_buy_in_exposure",
      "mcp__tools__compute_settlement_efficiency_kpi",
      "mcp__tools__run_pqc_timeline_fit",
      "mcp__tools__plan_tls_pki_migration",
      "mcp__tools__check_iso20022_pqc_readiness",
      "mcp__tools__check_fido_pqc_conformance",
      "mcp__tools__classify_blockchain_quantum_risk",
      "mcp__tools__run_sanctions_screening_fit",
      "mcp__tools__aggregate_ownership_50pct",
      "mcp__tools__score_fuzzy_match_calibration",
      "mcp__tools__classify_eccn_dual_use",
      "mcp__tools__assess_circumvention_diligence",
      "mcp__tools__build_no_russia_clause_pack",
      "mcp__tools__score_sanctions_screening_quality",
      "mcp__tools__run_mica_casp_fit",
      "mcp__tools__assess_mica_casp_readiness",
      "mcp__tools__calculate_mica_own_funds",
      "mcp__tools__lint_crypto_asset_whitepaper",
      "mcp__tools__assess_mar_crypto_surveillance",
      "mcp__tools__scope_mica_token_and_service",
      "mcp__tools__reconcile_mpp_subscription",
      "mcp__tools__model_tempo_gas_economics",
      "mcp__tools__score_partner_stablecoin_readiness",
      "mcp__tools__route_partner_stablecoin_jurisdiction",
      "mcp__tools__verify_dscsa_transaction_statement",
      "mcp__tools__verify_saleable_return",
      "mcp__tools__build_product_lineage",
      "mcp__tools__verify_product_authenticity",
      "mcp__tools__link_traceability_lot_code",
      "mcp__tools__anchor_document_integrity",
      "mcp__tools__verify_timestamp_attestation",
      "mcp__tools__verify_content_credential_signature",
      "mcp__tools__check_ai_act_art50_marking",
      "mcp__tools__verify_dual_layer_disclosure",
      "mcp__tools__verify_webbotauth_signature",
      "mcp__tools__audit_agent_key_rotation",
      "mcp__tools__crosswalk_agent_payment_rail_trust",
      "mcp__tools__assess_agent_directory_publish_readiness",
      "mcp__tools__verify_slsa_provenance",
      "mcp__tools__validate_openvex_statement",
      "mcp__tools__check_cra_annex1_completeness",
      "mcp__tools__assess_cra_vuln_reporting_readiness",
      "mcp__tools__classify_nis2_entity",
      "mcp__tools__check_nis2_art21_measures",
      "mcp__tools__calculate_nis2_penalty_exposure",
      "mcp__tools__score_nis2_incident_significance",
      "mcp__tools__score_nis2_supply_chain_diligence",
      "mcp__tools__check_nis2_governance_readiness",
      "mcp__tools__check_mcp_registry_entry",
      "mcp__tools__audit_mcp_tool_scope_revocation",
      "mcp__tools__check_emir_uti_completeness",
      "mcp__tools__reconcile_emir_pairing",
      "mcp__tools__run_emir_reporting_fit",
      "mcp__tools__assess_vida_drr_reporting_obligation",
      "mcp__tools__assess_vida_recapitulative_migration",
      "mcp__tools__classify_vida_platform_deemed_supplier",
      "mcp__tools__route_vida_oss_registration",
      "mcp__tools__run_vida_readiness_diagnostic",
      "mcp__tools__classify_eudr_commodity_scope",
      "mcp__tools__score_eudr_country_risk",
      "mcp__tools__link_eudr_supply_chain_traceability",
      "mcp__tools__run_eudr_readiness_fit",
      "mcp__tools__assess_iso42001_aims_conformance",
      "mcp__tools__classify_ai_system_governance",
      "mcp__tools__map_nist_ai_rmf_functions",
      "mcp__tools__check_gpai_code_conformance",
      "mcp__tools__run_ai_governance_fit",
      "mcp__tools__classify_ifrs17_measurement_model",
      "mcp__tools__check_ifrs17_risk_adjustment",
      "mcp__tools__calculate_solvency2_scr_ratio",
      "mcp__tools__reconcile_sii_ifrs17",
      "mcp__tools__run_insurance_reporting_fit",
      "mcp__tools__calculate_irrbb_eve_shocks",
      "mcp__tools__evaluate_irrbb_sot_eve",
      "mcp__tools__evaluate_irrbb_sot_nii",
      "mcp__tools__map_irrbb_standardised_approach",
      "mcp__tools__check_irrbb_csrbb_scope",
      "mcp__tools__run_irrbb_disclosure_fit",
      "mcp__tools__convert_markdown_document",
      "mcp__tools__convert_tabular_data",
      "mcp__tools__build_conversion_receipt",
      "mcp__tools__verify_conversion_receipt",
      "mcp__tools__prove_metadata_sanitization",
      "mcp__tools__build_digest_manifest",
      "mcp__tools__choose_cc_license",
      "mcp__tools__select_cbe_license",
      "mcp__tools__map_pil_flavor",
      "mcp__tools__certify_license_election",
      "mcp__tools__verify_license_election",
      "mcp__tools__generate_iscc_code",
      "mcp__tools__build_tdm_reservation",
      "mcp__tools__select_embedded_license",
      "mcp__tools__check_license_compatibility",
      "mcp__tools__assemble_license_terms",
      "mcp__tools__build_rights_record",
      "mcp__tools__generate_attribution_string",
      "mcp__tools__compute_ipfs_cid",
      "mcp__tools__compute_perp_margin",
      "mcp__tools__model_perp_position",
      "mcp__tools__compute_reg_z_appendix_j_apr",
      "mcp__tools__compute_trid_tolerance_cure",
      "mcp__tools__verify_trid_apr_accuracy",
      "mcp__tools__check_qm_points_and_fees",
      "mcp__tools__classify_qm_apr_apor_spread",
      "mcp__tools__compute_llpa_stack",
      "mcp__tools__check_agency_eligibility_matrix",
      "mcp__tools__check_conforming_loan_limit",
      "mcp__tools__compute_fha_mip_eligibility",
      "mcp__tools__compute_va_funding_fee_residual",
      "mcp__tools__lint_mismo_uldd_ulad",
      "mcp__tools__build_adverse_action_notice",
      "mcp__tools__compute_disparity_metrics",
      "mcp__tools__compute_hmda_rate_spread",
      "mcp__tools__compute_mla_mapr",
      "mcp__tools__compute_scra_rate_cap",
      "mcp__tools__check_card_act_ability_to_pay",
      "mcp__tools__test_hoepa_high_cost",
      "mcp__tools__test_hpml_escrow",
      "mcp__tools__classify_annex3_decisioning_obligations",
      "mcp__tools__test_bifsg_bias_thresholds",
      "mcp__tools__assess_naic_ais_program_readiness",
      "mcp__tools__lint_cbpr_structured_address",
      "mcp__tools__check_purpose_code_requirement",
      "mcp__tools__simulate_gpi_tracker_lifecycle",
      "mcp__tools__score_mt_mx_translation_fidelity",
      "mcp__tools__lint_lei_payment_binding",
      "mcp__tools__prevalidation_readiness_scorer",
      "mcp__tools__compute_remittance_disclosure",
      "mcp__tools__model_stablecoin_corridor_economics",
      "mcp__tools__compute_parametric_trigger_payout",
      "mcp__tools__run_illustration_selfsupport_test",
      "mcp__tools__compute_rbc_action_level",
      "mcp__tools__compute_lcm_rate_derivation",
      "mcp__tools__validate_openids_homeowners_record",
      "mcp__tools__calculate_claims_stp_economics",
      "mcp__tools__parse_camt053_reconciliation",
      "mcp__tools__compute_multilateral_netting",
      "mcp__tools__allocate_ihb_interest",
      "mcp__tools__test_hedge_effectiveness",
      "mcp__tools__score_cash_forecast_accuracy",
      "mcp__tools__amortize_asc606_commissions",
      "mcp__tools__reconcile_commission_statement",
      "mcp__tools__check_producer_license_reciprocity",
      "mcp__tools__compute_cdd_ownership_25pct",
      "mcp__tools__compute_perp_funding",
      "mcp__tools__assess_defi_lending",
      "mcp__tools__assess_restaking_risk",
      "mcp__tools__compute_pt_yt_yield",
      "mcp__tools__compile_work_mandate",
      "mcp__tools__check_genius_reserve_disclosure",
      "mcp__tools__assemble_mutual_nda",
      "mcp__tools__bind_agreement_acceptance",
      "mcp__tools__aggregate_reputation_score",
      "mcp__tools__verify_eth_state_proof",
      "mcp__tools__verify_reserve_proof",
      "mcp__tools__classify_sco60_exposure",
      "mcp__tools__optimize_social_security_claim_age",
      "mcp__tools__verify_acdc_delegation_chain",
      "mcp__tools__verify_anchored_extract",
      "mcp__tools__map_iso20022_to_evm_calldata",
      "mcp__tools__lint_besu_settlement_contract",
      "mcp__tools__check_linea_l2_finality_window",
      "mcp__tools__screen_onledger_transfer_batch",
      "mcp__tools__lint_settlement_orchestrator_conformance",
      "mcp__tools__verify_einvoice_vat_calc",
      "mcp__tools__route_einvoice_jurisdiction_mandate",
      "mcp__tools__build_einvoice_transmission_receipt",
      "mcp__tools__lint_compelling_evidence_ce30_agentic",
      "mcp__tools__compute_aca_affordability_safe_harbor",
      "mcp__tools__compute_esrp_exposure",
      "mcp__tools__build_226j_response_evidence_pack",
      "mcp__tools__run_section125_ndt",
      "mcp__tools__run_401k_adp_acp_test",
      "mcp__tools__lint_aiuc1_control_evidence",
      "mcp__tools__assemble_aiuc1_evidence_pack",
      "mcp__tools__lint_insurance_evidence_freshness",
      "mcp__tools__score_agent_insurability_evidence",
      "mcp__tools__build_claim_dispute_bundle",
      "mcp__tools__build_pld_disclosure_pack",
      "mcp__tools__derive_parametric_index_from_receipts",
      "mcp__tools__assess_traiga_exposure",
      "mcp__tools__build_traiga_safe_harbor_pack",
      "mcp__tools__lint_ab2013_training_data_disclosure",
      "mcp__tools__check_sb53_frontier_scope",
      "mcp__tools__reconcile_erc8056_multiplier",
      "mcp__tools__map_robinhood_chain_regime",
      "mcp__tools__lint_stock_token_valuation",
      "mcp__tools__compute_stock_token_collateral_haircut",
      "mcp__tools__classify_bold_challenge_finality",
      "mcp__tools__stress_test_ap_redemption_path",
      "mcp__tools__run_robinhood_chain_fit_diagnostic",
      "mcp__tools__compute_npv",
      "mcp__tools__compute_irr",
      "mcp__tools__compute_xirr",
      "mcp__tools__compute_annuity",
      "mcp__tools__compute_breakeven",
      "mcp__tools__compute_bond_duration",
      "mcp__tools__compute_dv01",
      "mcp__tools__compute_convexity",
      "mcp__tools__build_amortization_schedule",
      "mcp__tools__compute_dti_ratios",
      "mcp__tools__compute_ltv_ratios",
      "mcp__tools__compute_federal_withholding",
      "mcp__tools__compute_gross_to_net",
      "mcp__tools__compute_flsa_regular_rate",
      "mcp__tools__compute_escrow_analysis",
      "mcp__tools__compute_mlr_rebate",
      "mcp__tools__compute_experience_mod",
      "mcp__tools__score_credit_model_quantized",
      "mcp__tools__lint_fedwire_structured_address",
      "mcp__tools__sweep_fedwire_addresses",
      "mcp__tools__check_etr_control_evidence",
      "mcp__tools__build_etr_possession_chain",
      "mcp__tools__compute_rwa_erba_2026",
      "mcp__tools__compute_oprisk_sma_2026",
      "mcp__tools__simulate_output_floor",
      "mcp__tools__build_idv_session_receipt",
      "mcp__tools__check_camera_provenance",
      "mcp__tools__compute_raroc_loan_price",
      "mcp__tools__compute_dscr",
      "mcp__tools__compute_lcr_nsfr_leverage",
      "mcp__tools__compute_globe_topup_tax",
      "mcp__tools__price_embedded_insurance",
      "mcp__tools__compute_cross_border_fees",
      "mcp__tools__compute_fx_netting_positions",
      "mcp__tools__run_rate_shock_ladder",
      "mcp__tools__replay_supervisory_scenario",
      "mcp__tools__simulate_var_monte_carlo",
      "mcp__tools__recompute_fund_nav",
      "mcp__tools__test_nav_error_materiality",
      "mcp__tools__compute_fund_expense_ratios",
      "mcp__tools__score_payee_name_match",
      "mcp__tools__build_vop_session_receipt",
      "mcp__tools__build_agent_test_evidence",
      "mcp__tools__build_agent_incident_record",
      "mcp__tools__build_ai_workpaper_record",
      "mcp__tools__check_agent_token_scope",
      "mcp__tools__lint_cbom_structure",
      "mcp__tools__compute_pqc_deadline_ladder",
      "mcp__tools__convert_tempo_fee_amm",
      "mcp__tools__compute_tempo_mainnet_fee_capacity",
      "mcp__tools__compute_canton_traffic_cost",
      "mcp__tools__compute_canton_app_reward_estimate",
      "mcp__tools__lint_x402_v2_migration",
      "mcp__tools__compute_15c3_3_reserve",
      "mcp__tools__lint_trace_cat_reports",
      "mcp__tools__lint_metro2_record",
      "mcp__tools__lint_x12_claim_records",
      "mcp__tools__check_official_statement_completeness",
      "mcp__tools__check_debt_validation_notice",
      "mcp__tools__check_retail_installment_disclosures",
      "mcp__tools__check_private_student_loan_disclosures",
      "mcp__tools__estimate_cross_venue_margin_capital",
      "mcp__tools__run_umr_aana_readiness",
      "mcp__tools__assemble_ocg_evidence_bundle",
      "mcp__tools__check_dpa_gdpr_art28",
      "mcp__tools__score_clause_coverage",
      "mcp__tools__assemble_ai_addendum",
      "mcp__tools__map_ai_act_procurement_clauses",
      "mcp__tools__screen_sanctions_private",
      "mcp__tools__compute_rbc_action_level_private",
      "mcp__tools__check_capital_adequacy_private",
      "mcp__tools__build_idv_verification_incident_record",
      "mcp__tools__verify_witness_cosignatures",
      "mcp__tools__compute_large_exposures_limit",
      "mcp__tools__calculate_cecl_ecl_allowance",
      "mcp__tools__compute_discount_window_capacity",
      "mcp__tools__compute_cyber_incident_notification_clock",
      "mcp__tools__compute_var_backtest_traffic_light",
      "mcp__tools__compute_fdic_assessment_rate",
      "mcp__tools__map_call_report_schedule_rc",
      "mcp__tools__map_call_report_schedule_rcr",
      "mcp__tools__run_call_report_edit_checks",
      "mcp__tools__map_bhc_schedule_hc",
      "mcp__tools__map_bhc_schedule_hcr",
      "mcp__tools__compute_fr2052a_inflow_outflow_classification",
      "mcp__tools__compose_eval_attestation_receipt",
      "mcp__tools__rollforward_y14_capital_worksheet",
      "mcp__tools__map_nmd_behavioral_repricing",
      "mcp__tools__calculate_basis_risk_nii_shock",
      "mcp__tools__compute_basel_haircut_adjusted_exposure",
      "mcp__tools__check_credit_concentration_topn_sector",
      "mcp__tools__compute_counterparty_limit_check",
      "mcp__tools__check_securitization_risk_retention",
      "mcp__tools__track_ifrs17_loss_component_rollforward",
      "mcp__tools__aggregate_solvency2_scr_modules",
      "mcp__tools__build_model_inventory_entry",
      "mcp__tools__build_ai_training_data_lineage_record"
    ],
    "deny": []
  }
}

Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. It assumes the server is installed as tools. Review it before you use it.

Keep track

Get told when this server's tools change: a new tool, a new description, a new class.

https://anlyon.dev/servers/co.ainumbers/tools/changes.xml

An Atom feed. Works in any feed reader or Slack's RSS app.

Anlyon alternative

402 of these tools change something. Run those as governed actions.

Declare what an action changes, and Anlyon puts it under a shared limit, binds every approval to the exact request, and grades the outcome. It works on any HTTPS API you can describe as an action.

Anlyon does not run, proxy or govern this MCP server. A tool your own code calls directly is not routed through Anlyon and is not governed by it.

Import an OpenAPI spec →

Compare with other servers for the same product.

Source: MCP Registry · repository. Synced Oct 21, 2026, 17:46 UTC.

For agents: Markdown · JSON · report this listing