DOCS /

documentation

Everything you need to integrate con+ into your agent.

npx conplus-mcpclick to copy

Add to your MCP config to give any compatible AI agent token risk scoring capabilities as a native tool call.

MCP CONFIG (CLAUDE_DESKTOP_CONFIG.JSON)

{
  "mcpServers": {
    "conplus": {
      "command": "npx",
      "args": ["conplus-mcp"],
      "env": {
        "CONPLUS_API_URL": "https://conplus.app",
        "CONPLUS_CHAIN": "base"
      }
    }
  }
}

Available Tools

conplus_analyze

Score a single pair address on Base. Returns tier, risk_score, risk_24h, risk_7d, risk_factors, trust_signals, and AI narrative.

Example Tool Call

// Agent calls conplus_analyze
{
  "tool": "conplus_analyze",
  "input": {
    "pair_address": "0x3F4Bc91A...",
    "chain": "base"
  }
}

// con+ returns
{
  "tier": "CRITICAL",
  "risk_score": 0.921,
  "warning_24h": 0.921,
  "warning_7d": 0.961,
  "risk_summary": "This token exhibits very high risk patterns. Multiple rug pull indicators detected.",
  "recommendation": "Avoid all interaction with this token."
}