Live alarm claims in the caller's scope (claims-only read)
What a console refetches on the claims_changed stream message instead of the whole queue. The live claims (operator_id, operator_name, claimed_at, expires_at, expires_in_seconds) on queued alarms in the caller's tenant scope, keyed by event id, with the queue's own membership predicate. An alarm with no live claim is absent, so for the alarms the request covers absence means free. Optional site_id and repeated event_ids (at most 500) narrow what is covered. Central staff only: customer-side callers get an empty map and never staff identity. activity (claim-lock R2) is a sibling map of escalation activity per event id (a valid intent and/or live calls, derived from the session records); it is NOT inside claims because activity outlives the claim. A keyed read (event_ids) covers every queued alarm named, an unfiltered or site_id-only read the alarms with a live claim; empty while CLAIM_ESCALATION_ENABLED is off. Requires claim participation (review_label_alarms, or, for central staff, any of escalate_voice_down, escalate_dispatch, notify_property_owner: whoever the claim refusal can refuse). Malformed ids are 404; also 404 while CLAIM_CLEANUP_ENABLED is off (read claim_cleanup_enabled on GET /auth/me first).
Request
- Base URLhttps://api.verdun.ai (the document declares no server; this is the origin it was published from)
- URLhttps://api.verdun.ai/events/claims
- AuthHTTP bearer
Query parameters
Only alarms at this site.
Only these alarms (repeat the parameter); at most 500 IDs to bound a single explicit filter. This cap does not limit an unfiltered response's size.
Response
Successful Response
Changes
No recorded changes to this operation across all 2 revisions of this API.