Guard prompt before model execution

POST 
/api/v1/guard

Returns allow/block verdict plus tracking IDs.

Request

Responses

200

Guard verdict (allow, fail-open fallback, or certain strict dependency blocks).

400

Invalid payload

401

Unauthorized (missing/invalid API key or system mismatch)

403

Blocked by policy/risk

413

Payload too large

429

Rate limited

Response Headers

Retry-After
X-RateLimit-Limit
X-RateLimit-Remaining

503

Infrastructure/configuration failure (strict mode paths, encryption key missing, etc.)