MCP Tool Reference 🟡 Beta
scan_token
Input: { "contract": "<address>", "chain": "solana" | "ethereum" | "base" | ... }
Output: full RMI risk report — checks run, evidence, confidence, verdict.
Identical schema to POST /scan on the REST API (see Scanner API).
quick_verdict
Input: { "contract": "<address>" } → Output: { "verdict": "PASS|CAUTION|AVOID", "top_risks": [...] }
wallet_profile
Input: { "address": "<address>", "chain": "<chain>" } → Output: behavior
profile, counterparty risk flags, funding chain depth.
security_context
Input: { "query": "<natural language>", "top_k": 5 } → Output: ranked
security patterns with sources (backed by the open RugBrain corpus).
Error model
| Code | Meaning |
|---|---|
| 402 | Payment required — response carries x402 payment instructions |
| 403 | Payment received but failed facilitator verification |
| 422 | Bad input (invalid address, unsupported chain) |
| 429 | Rate limited |