Agent Discovery

How AI agents discover and use verified.md endpoints.

Discovery Endpoint

The primary discovery endpoint is:

https://verified.md/agents.md

This returns a structured manifest that AI agents can parse to understand all available capabilities.

Manifest Structure

# verified.md Agent Discovery

## Available Endpoints
- /verify/{domain} - Primary verification
- /verify/{domain}?signals=true - Full trust scoring
- /lookup/{entity-id} - Reverse lookup

## Documentation Endpoints
- /agents.md - This manifest
- /api.md - API reference
- /mcp.md - MCP configuration

Integration Pattern

1. Agent reads /agents.md at startup
2. Agent parses available endpoints
3. Before any transaction:
   - Call /verify/{domain}
   - Check response.data.verified
   - If false, flag for human review

Trust Signal Response

When using ?signals=true, agents receive actionable guidance:

{
  "agent_guidance": {
    "action": "PROCEED",
    "safe_actions": ["payment", "data_sharing"],
    "max_transaction_usd": 10000,
    "requires_human": false
  }
}