Evaluate a conversation for an agent loop
Run the deterministic loop detector over the conversation’s recent message window and return its evaluation — whether the conversation is stuck in a loop, the reason codes, and the customer-turn count behind the verdict. Every body field is optional and tunes the detector per request (window size, similarity threshold, sentiment threshold, minimum customer turns); omitting the body uses the calibrated defaults. Read-only — it evaluates without mutating the conversation or triggering a handoff.
Authorizations
Dashboard JWT token from Clerk
Headers
Stripe-style idempotency token. Pass a stable, client-generated value (1-255 chars) to dedupe retries on transient timeouts. The same key+credential+path replays the original response for 24h on 2xx (5min on 4xx, 30s on 5xx). Returns 409 if a concurrent request with the same key is already in flight; replayed responses include the Idempotency-Replay: true response header.
1 - 255Sandbox opt-in for Clerk-session-authenticated requests. Set to true to route the call through the test-mode pipeline: no real provider delivery, no credits deducted, response meta.test_mode: true. Ignored for live API keys (dv_live_sk_*) — server-to-server clients must use a test-prefixed key (dv_test_sk_*) to exercise sandbox. Test-prefixed keys unconditionally enable sandbox regardless of this header.
true, false Path Parameters
The agent conversation id.
1 - 200Body
Optional detector tuning — omit for the calibrated defaults.
Optional detector tuning — omit for the calibrated defaults.
How many recent turns to inspect (default 5).
2 <= x <= 20Minimum re-ask similarity that counts as a repeated demand.
0 <= x <= 1Customer-turn sentiment below this counts toward a frustration loop.
-1 <= x <= 1Minimum inbound turns before a loop verdict can fire.
1 <= x <= 20