Zarq Risk Intelligence

Smithery

Independent crypto risk intelligence via the Model Context Protocol (MCP). Trust Score ratings, Distance-to-Default analysis, structural collapse warnings, and pre-trade safety checks for 198 tokens.

Quick Start

Remote (SSE / Streamable HTTP)

No installation needed — connect directly:

{
  "mcpServers": {
    "zarq-crypto": {
      "url": "https://mcp.zarq.ai/mcp"
    }
  }
}

Local (stdio)

pip install mcp httpx
python zarq_mcp_server.py

Tools (8)

Tool	Description
crypto_safety_check	Quick pre-trade safety validation (<100ms)
crypto_rating	Full Trust Score with 5-pillar breakdown (A+ to F)
crypto_dtd	Distance-to-Default with 7 distress signals
crypto_signals	Active Structural Collapse/Stress signals
crypto_compare	Head-to-head token risk comparison
crypto_distress_watch	All tokens with DtD < 2.0
crypto_alerts	Structural collapse/stress warnings
crypto_ratings_bulk	Bulk ratings for multiple tokens

Performance

• 100% death recall — every token collapse detected (out-of-sample)
• 98% precision — only 1 genuine false positive in OOS validation
• 198 tokens rated with daily updates
• <100ms response time for safety checks

Architecture

Client (Claude, Cursor, etc.)
    ↓ MCP Protocol
ZARQ MCP Server
    ↓ HTTPS
zarq.ai API (free, no key required)

API

Free API at zarq.ai. No API key required during Early Access.

Tags

crypto risk defi safety trust-score crash-prediction distance-to-default ratings blockchain token-analysis

License

MIT