layout: default title: API Reference

API Reference

The SecAI Radar API allows clients (browsers, scripts, and autonomous AI agents) to query the current trust rankings and specific integration details.

All API routes are served under the /api/v1/public prefix.

Endpoints

1. GET /mcp/rankings

Returns the global rankings for Model Context Protocol (MCP) servers.

Response Structure:

{
  "attestation": {
    "assessedBy": "SecAI Radar",
    "methodologyVersion": "v1.0"
  },
  "data": {
    "items": [
      {
        "mcpId": "srv-123",
        "name": "Example MCP",
        "provider": "VendorX",
        "tier": "A",
        "trustScore": 92.5
      }
    ]
  },
  "meta": {
    "total": 150,
    "page": 1,
    "pageSize": 20
  }
}

2. GET /agents/rankings

Returns the global rankings for standalone AI Agents. The response structure maps identically to the MCP rankings.

3. POST /submissions

Allows integrations to manually register for scouting.

Request Payload:

{
  "repoUrl": "https://github.com/organization/my-new-mcp",
  "integrationType": "mcp", 
  "contactEmail": "dev@example.com"
}

(Note: integrationType accepts "mcp" or "agent")

For a full interactive OpenAPI specification, boot the API locally and navigate to http://localhost:8000/docs.