Requeue Failed Queue Job
Requeue a terminally failed or cancelled job for one more execution (raises its attempt budget by one). The operator remedy for crashed runs under the no-silent-re-execution default. Requeueing a CANCELLED job whose cancellation intent is still live requires clear_cancellation=true - an explicit override of the recorded cancel; without it the re-driven attempt is re-cancelled at its first checkpoint. Requeueing a job that FAILED within the last lock_grace seconds requires force=true: its worker may still be executing (swept-but-alive), and a second producer would race unfenced writes.
Authorizations
Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
Path Parameters
Response
Successful Response
One durable queue job (a background run's queue ticket).
Job ID (identical to the run's run_id)
Component kind: agent, team or workflow
ID of the agent/team/workflow the run executes on
Session the run belongs to
queued | running | completed | failed | cancelled | paused
Kind of work the ticket carries (runs only today)
User the run was submitted for
Serialized run parameters
Executions started so far
Execution budget under any failure mode
Client-provided dedupe key, if any
Epoch seconds the job becomes claimable
Worker currently holding the job's lock
Epoch seconds the current lock was taken
Terminal error of the last attempt, if any
Epoch seconds the job was accepted
Epoch seconds of the last state change
Epoch seconds the job reached a terminal state