History

cryptorugmunch cfd75fd1a0 fix(lint): drop ruff errors from 1470 to 0 across app/ and tests/ - Exclude generated SDK (sdks/python) and operational scripts from ruff lint - Add targeted per-file ignores for ASYNC*, S310, S603, S607, S108, S314, S102, PIE810, SIM102 in scripts/ - Auto-fix safe categories (I001, F401, W292, F841, PIE790, RUF100, etc.) - Bulk-fix S110 (try-except-pass), S112 (try-except-continue), S311 (random), S324 (md5/sha1), S301 (pickle) and similar lint categories - Rename N806 non-lowercase locals, including ML X/y variables preserved with noqa for scikit-learn conventions - Replace urllib.request calls with httpx.AsyncClient / httpx.Client (S310) - Wrap blocking os.path/os calls in asyncio.get_running_loop().run_in_executor - Replace subprocess.run with asyncio.create_subprocess_exec in async contexts - Store asyncio.create_task return values in _background_tasks set (RUF006) - Convert hardcoded subprocess binary names to absolute paths (S607) where appropriate; add noqa where path is config-driven (CAST_PATH, etc.) - Parameterize SQL queries with placeholders and add noqa for sanitized inputs - Fix all mechanical categories: SIM102, PIE810, TC001/2/3, S108, S314, S107, S306, S301, N802/N815/N817, S104, S605, S501, RUF022, UP031 - Add missing 'import asyncio' where referenced but not imported (F821) - Fix E402 module-import-not-at-top by adding '# noqa: E402' for circular-import safe cases and code-defined imports - Remove hardcoded Redis password in databus_warm_cron.py; use env vars Tests: - Add tests/unit/core/test_ai_router.py (8 tests): model resolution, chat completion with mocked httpx, fallback to OpenRouter, no-provider error, streaming - Add tests/unit/core/test_tracing.py (7 tests): setup_otel disabled/enabled, shutdown_otel, span helpers, tracing-enabled route registration - Add tests/unit/core/test_langfuse.py (2 tests): no-env init, noop flush - Fix tests/unit/domain/scanner/test_service.py to import from the moved app.domains.scanners.core.service Result: 'ruff check .' passes with 0 errors (was 1470). Pytest: 808 passed, 1 skipped (no regressions).		2026-07-08 03:36:58 +07:00
..
python	fix(lint): drop ruff errors from 1470 to 0 across app/ and tests/	2026-07-08 03:36:58 +07:00
typescript	merge: chore/cleanup-remove-bloat-and-secrets into main	2026-07-02 01:24:22 +07:00
README.md	merge: chore/cleanup-remove-bloat-and-secrets into main	2026-07-02 01:24:22 +07:00
smithery.yaml	merge: chore/cleanup-remove-bloat-and-secrets into main	2026-07-02 01:24:22 +07:00

README.md

Rug Munch Intelligence — MCP Server

MCP server exposing crypto intelligence tools to AI agents.

8 tools for AI agents. Real-time token risk, wallet analysis, deployer reputation, news sentiment, full AI research reports, RAG search, similar tokens, and cross-chain entity resolution. 13+ chains (Solana, Ethereum, Base, Arbitrum, Optimism, Polygon, BSC, Tron, Bitcoin, Avalanche, Fantom, Gnosis + EVM subnets). x402 paid tier.

Quick start

Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "rugmunch": {
      "url": "https://mcp.rugmunch.io/mcp",
      "transport": "streamable-http"
    }
  }
}

Cursor

Add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "rugmunch": {
      "url": "https://mcp.rugmunch.io/mcp",
      "transport": "streamable-http"
    }
  }
}

Continue.dev

Add to ~/.continue/config.json:

{
  "experimental": {
    "modelContextProtocolServers": [
      {
        "name": "rugmunch",
        "url": "https://mcp.rugmunch.io/mcp"
      }
    ]
  }
}

Available tools

Tool	Description	x402 price
`get_token_risk`	Real-time risk score for any token across 13+ chains	FREE 5/day, $0.01
`get_wallet_analysis`	Wallet activity, balance, history, reputation	FREE 5/day, $0.01
`get_deployer_reputation`	Deployer reputation 0-100, deterministic	$0.02
`get_news_sentiment`	Latest news + composite sentiment	FREE 5/day, $0.01
`generate_report`	7-section LLM research report	$5.00
`query_catalog`	Natural language catalog query	$0.05
`find_similar_tokens`	Vector-similar tokens	$0.03
`resolve_entity`	Cross-chain entity resolution	$0.10

Authentication

Free tier: 5 calls/day, no auth. Paid tier: include X-Payment: <base64(tx_hash+signature)> header.

Architecture

OpenAPI: https://api.rugmunch.io/openapi.json
MCP server: https://api.rugmunch.io/mcp (JSON-RPC 2.0 + plain JSON)
x402 catalog: https://api.rugmunch.io/api/v1/x402/catalog
Health: https://api.rugmunch.io/health

Direct tool calls (no MCP)

# Get token risk
curl -X POST https://api.rugmunch.io/mcp/call/get_token_risk \
  -H "Content-Type: application/json" \
  -d '{"arguments":{"chain":"solana","address":"DezXAZ..."}}'

# Generate AI report
curl -X POST https://api.rugmunch.io/mcp/call/generate_report \
  -H "Content-Type: application/json" \
  -d '{"arguments":{"subject_type":"token","subject_id":"solana:DezXAZ..."}}'

Python SDK

pip install rugmunch

import asyncio
from rugmunch_sdk import RMI

async def main():
    async with RMI(base_url="https://api.rugmunch.io") as client:
        risk = await client.get_token_risk("solana", "DezXAZ...")
        print(f"Risk: {risk.score}/100 ({risk.tier})")
        report = await client.generate_report("token", "solana:DezXAZ...")
        print(report.markdown[:500])

asyncio.run(main())

TypeScript SDK

npm install @rugmunch/sdk

import { RMI } from "@rugmunch/sdk";
const client = new RMI({ baseUrl: "https://api.rugmunch.io" });
const risk = await client.getTokenRisk("solana", "DezXAZ...");