PLATFORM / AGENT IDENTITY

Every agent has an identity

Client credentials, authorization scopes, and risk classification. No agent calls Meerkat without being registered and authenticated first.

How It Works

Register, authenticate, govern

Register

Every agent gets a unique agent ID, client credentials, and authorized domains. Registration happens through the Console or the API.

Authenticate

OAuth 2.0 client credentials flow. The agent presents its credentials, receives a scoped token, and uses it for all subsequent API calls.

Govern

Authorization scopes control what each agent can do. Heartbeat monitoring confirms it is alive. The audit trail logs everything.

Register an agent
          terminal
          
        
# Register a new AI agent
curl -X POST https://api.meerkatplatform.com/v1/agents \
  -H "Authorization: Bearer $MEERKAT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Clinical Scribe",
    "domain": "healthcare",
    "risk_classification": "high"
  }'
Exchange credentials for a token
          terminal
          
        
# Get an agent access token
curl -X POST https://api.meerkatplatform.com/v1/agents/token \
  -H "Content-Type: application/json" \
  -d '{
    "client_id": "agt_healthcare_clinical_scribe_a1b2c3",
    "client_secret": "ags_live_xxxxxxxxxxxxxxxx"
  }'

Register your first agent

Free tier. 10,000 verifications/month. No credit card.

Get Started Read the Docs