Skip to content

Trust Check

by gmahar82-stack · remote · no sign-in to list tools · v1.1.0 · published by GitHub user gmahar82-stack

Base crypto safety for agents: honeypot/tax test, address poisoning, rug feed, creator record.

Changes · 5
CompareRegistry entry

5 tools: 5 change things.

Scanned Oct 21, 2026

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

claude mcp add --transport http trust-check https://trust-check.gm-tools.workers.dev/mcp

Tools

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

5 tools

Changes things 5Claude Code: ask
  • check_address_safetyBefore sending crypto on Base, check the destination: scam blacklists, address poisoning (a look-alike of an address you really use; pass `from` = your wallet), token contracts and…No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Before sending crypto on Base, check the destination: scam blacklists, address poisoning (a look-alike of an address you really use; pass `from` = your wallet), token contracts and burn addresses where funds are lost, never-used addresses. Returns a verdict with reasons. 5 free checks per day, then $0.002 per check via x402 (USDC on Base).

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: true
  • check_token_creatorBefore buying a newly launched Base token, check the wallet that deployed it: how many of its earlier tokens Rug Watch followed, and how many collapsed (rug pull or dump).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Before buying a newly launched Base token, check the wallet that deployed it: how many of its earlier tokens Rug Watch followed, and how many collapsed (rug pull or dump). Verdict: serial_collapser / collapsed_before / no_collapses_seen / unknown. 5 free calls per day, then $0.002 per call via x402 (USDC on Base).

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: true
  • check_token_safetyBefore buying a token on Base, check it: simulates a real buy and sell (honeypot and hidden-tax test), reads owner powers (mint, blacklist, fee changes, pause), whale concentration…No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Before buying a token on Base, check it: simulates a real buy and sell (honeypot and hidden-tax test), reads owner powers (mint, blacklist, fee changes, pause), whale concentration, liquidity lock and scam lists. Returns a verdict (scam / high_risk / caution / low_risk) with plain-language reasons. 5 free checks per day, then $0.002 per check via x402 (USDC on Base).

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: true
  • deep_token_safety_reportFull due diligence before a bigger buy of a Base token: the token check (buy/sell simulation, taxes, owner powers, whales, liquidity, scam lists) plus the creator's Rug Watch recor…No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Full due diligence before a bigger buy of a Base token: the token check (buy/sell simulation, taxes, owner powers, whales, liquidity, scam lists) plus the creator's Rug Watch record, wallet check and first funder, the owner wallet, the token's liquidity history, website domain age, how often new Base tokens collapse, an AI summary and a what-to-do list. $2 per report via x402 (USDC on Base), no free calls; failures are not charged.

    Annotations the server declares

    • readOnlyHint: true
    • openWorldHint: true
  • recent_liquidity_collapsesNew Base tokens whose pool liquidity collapsed (fell 90%+ from its peak: a rug pull or a heavy dump) in the last hours, newest first, with peak and remaining liquidity, hours from …No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    New Base tokens whose pool liquidity collapsed (fell 90%+ from its peak: a rug pull or a heavy dump) in the last hours, newest first, with peak and remaining liquidity, hours from launch and the creator wallet. Use it to exit positions and blocklist serial creators. 5 free calls per day, then $0.002 per call via x402 (USDC on Base).

    Annotations the server declares

    • readOnlyHint: 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.

Install

Pinned to version 1.1.0. Values in angle brackets are yours to fill in.

claude mcp add --transport http trust-check https://trust-check.gm-tools.workers.dev/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": [],
    "ask": [
      "mcp__trust-check__check_token_safety",
      "mcp__trust-check__check_address_safety",
      "mcp__trust-check__recent_liquidity_collapses",
      "mcp__trust-check__check_token_creator",
      "mcp__trust-check__deep_token_safety_report"
    ],
    "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 trust-check. Review it before you use it.