Skip to content

menantonio83-hue/tnt-house-risk-data-api

v1.14.0MIT

Solana token risk-scoring for AI trading agents. One call returns a 0-100 safety score, on-chain-provable insider wallet cluster detection (shared first-funder tracing), mint/freeze authority status, honeypot risk, and LP-lock status for any Solana token mint. Works with zero setup: 3 free anonymous calls/day, no signup, no API key. x402 pay-per-call ($0.02/call in USDC) is also supported for fully autonomous agents. Free key with a 15/day quota at tnt-audit.com/risk-api.

GitHubOpen Website

TNT House Risk-Data API — Agent Plugins package

This directory is an Agent Plugins 1.0.0 package for the Risk-Data API's MCP server. It lets Agent Plugins-compatible clients (Codex, Claude Code, Cursor, and others) discover and connect to the same production MCP server already listed on Smithery, Glama, and the Official MCP Registry — no separate integration needed.

  • plugin.json — manifest (name, description, repo, license, keywords)
  • mcp.json — declares the remote server at https://tnt-audit.com/api/mcp (streamable-http transport)

What you get with zero configuration

  • check_token_risk works with no API key for the first 3 calls/day per IP (100/day global cap) — a genuine anonymous trial, not a locked demo.
  • Returns a 0-100 safety score, insider wallet cluster detection (shared first-funder tracing), mint/freeze authority status, honeypot risk, and LP-lock status for any Solana token mint.

Higher limits / other tools

check_token_risk_batch (up to 25 mints/call) and get_token_risk_history require a Bearer API key. Per the Agent Plugins spec, mcp.json cannot embed credentials in headers — get a free key (15 calls/day) at tnt-audit.com/risk-api and add it as an Authorization: Bearer <key> header wherever your client exposes that option for configured MCP servers.

x402 pay-per-call ($0.02/call in USDC, no key at all) is also available for fully autonomous agents — see tnt-audit.com/risk-api for the x402 endpoint.

Links