KushBitx AgentProof

by kushbitx.com · remote · no sign-in to list tools · v1.0.0 · published by com.kushbitx

Safety checks for AI agents paying on Base: free SpendGuard, paid x402 token risk and preflight.

Final · 3Changes · 2

Compare Registry entry

5 tools: 3 do something that can't be undone and 2 change things.

Scanned Oct 22, 2026

Deletes or cancels0Claude Code: deny
Can't be undone3Claude Code: ask: see these tools
Changes things2Claude Code: ask: see these tools
Only reads0Claude Code: allow

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

claude mcp add --transport http agentproof https://kushbitx.com/mcp

Findings

Matches of fixed checks, with the text they matched.

3 warn

Warnspendguard_evaluate
The server marks "spendguard_evaluate" 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":true}; The name "spendguard_evaluate" says "evaluate": a word for a change.
What this check looks for
Warnpreflight
The server marks "preflight" 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":true}; The input addresses people (input.to): what it sends cannot be taken back.
What this check looks for
Warnverify_payment
The server marks "verify_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":true}; The name "verify_payment" says "payment": a word that usually means money moves, which cannot be taken back.
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.

5 tools

Can't be undone 3Sends, pays, publishes or merges: once done, it cannot be taken back.Claude Code: ask

preflightSimulate an unsigned Base transaction and detect failed execution, dangerous approvals and destination risk signals.Reads text others writeCan send data out1 finding
Why it is in this group
- The input addresses people (input.to): what it sends cannot be taken back.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
- outboundThe input takes a destination (input.to): what the tool writes goes there.
Findings
- Warn
  The server marks "preflight" 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":true}; The input addresses people (input.to): what it sends cannot be taken back.
  What this check looks for
Description, as the server wrote it
Simulate an unsigned Base transaction and detect failed execution, dangerous approvals and destination risk signals. Paid tool: 0.00098 USDC per call on Base via x402 (exact scheme). An unpaid call returns the payment requirement and charges nothing.
Annotations the server declares
- title: Transaction Preflight
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: true
spendguard_evaluateFree.Reads text others writeCan send data out1 finding
Why it is in this group
- The name "spendguard_evaluate" says "evaluate": a word for a change.
- The input addresses people (input.recipient): what it sends cannot be taken back.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
- outboundThe input takes a destination (input.recipient): what the tool writes goes there.
Findings
- Warn
  The server marks "spendguard_evaluate" 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":true}; The name "spendguard_evaluate" says "evaluate": a word for a change.
  What this check looks for
Description, as the server wrote it
Free. Check a proposed Base USDC payment against a spending policy you supply (per-transaction limit, daily budget, human-approval threshold, allowed recipients) before money moves. Returns APPROVE, HUMAN_APPROVAL or BLOCK with findings. Advisory only: it never signs, holds keys or moves funds.
Annotations the server declares
- title: SpendGuard policy check
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: true
verify_paymentVerify an onchain Base USDC payment, its recipient, amount, status and confirmations.Reads text others write1 finding
Why it is in this group
- The name "verify_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.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Findings
- Warn
  The server marks "verify_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":true}; The name "verify_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
Verify an onchain Base USDC payment, its recipient, amount, status and confirmations. Paid tool: 0.00288 USDC per call on Base via x402 (exact scheme). An unpaid call returns the payment requirement and charges nothing.
Annotations the server declares
- title: Payment Proof
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: true

Changes things 2Creates or changes something that can be changed back.Claude Code: ask

quick_checkA deterministic Base chain identity and official USDC contract-presence check.Reads text others writeNo findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Description, as the server wrote it
A deterministic Base chain identity and official USDC contract-presence check. No request body. Not a token-risk report, transaction simulation, or payment proof. Paid tool: 0.00098 USDC per call on Base via x402 (exact scheme). An unpaid call returns the payment requirement and charges nothing.
Annotations the server declares
- title: AgentProof Quick Check
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: true
token_riskScreen a Base token for sellability, taxes, privileged controls, holder concentration and market risk.Reads text others writeNo findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Description, as the server wrote it
Screen a Base token for sellability, taxes, privileged controls, holder concentration and market risk. Paid tool: 0.00098 USDC per call on Base via x402 (exact scheme). An unpaid call returns the payment requirement and charges nothing.
Annotations the server declares
- title: Token Risk
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: true

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.

Exfil surface

Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.

No private-data, untrusted-input, outbound path in this server alone: no tool is tagged private-data.

Reads private data0: Reads files, mail, messages, databases, secrets, calendars or repositories.
Reads text others write5: Brings in text someone else can write: web pages, inbound email, messages, issues and comments.; preflight
quick_check
spendguard_evaluate
token_risk
verify_payment
Can send data out2: Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.; preflight
spendguard_evaluate

Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.

Install

Pinned to version 1.0.0. Values in angle brackets are yours to fill in.

claude mcp add --transport http agentproof https://kushbitx.com/mcp

Run in a terminal.

Claude Code permissions

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

.claude/settings.json

{
  "permissions": {
    "allow": [],
    "ask": [
      "mcp__agentproof__spendguard_evaluate",
      "mcp__agentproof__quick_check",
      "mcp__agentproof__token_risk",
      "mcp__agentproof__preflight",
      "mcp__agentproof__verify_payment"
    ],
    "deny": []
  }
}

Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as agentproof. Review it before you use it.

Capability tags behind 5 rules

mcp__agentproof__spendguard_evaluate untrusted-input, outbound
mcp__agentproof__quick_check untrusted-input
mcp__agentproof__token_risk untrusted-input
mcp__agentproof__preflight untrusted-input, outbound
mcp__agentproof__verify_payment untrusted-input

Keep track

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

https://anlyon.dev/servers/com.kushbitx/agentproof/changes.xml

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

Anlyon alternative

5 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. Synced Oct 22, 2026, 14:17 UTC.

For agents: Markdown · JSON · report this listing