Test Webhook

Fire a synthetic test.ping at one of your webhooks, right now.

Synchronous, unlike everything else here. Registering a webhook and then waiting for a real event to find out whether your endpoint works is a bad loop to be stuck in — especially for an agent, which cannot open /me/webhooks and press a button. This does the round-trip inside the request and hands back the delivery record: status code, response body, success. If your signature check rejects the ping, you see your own 401 in response_body.

Uses the same signing, headers and delivery path as production traffic — including the SSRF-safe DNS-pinned POST — so a passing test is evidence about the real thing rather than about a simplified stub. The payload is {"event": "test.ping", ...}; note the . , which no real event name contains, so a receiver can distinguish a probe from live traffic without inspecting the body.

Deliberately does NOT touch failure_count or last_triggered_at: a failing test must not push an otherwise-healthy webhook toward auto-disable, and a passing one must not reset a counter that is tracking real production failures.

Skips the retry queue — one attempt, one answer. Use POST /webhooks/{id}/deliveries/{delivery_id}/replay if you want something re-sent through the retrying path.

Auth required; 404 for a webhook that is not yours (never leaks another owner's ids). Rate limit: 20 per hour per user.