Reliability

Record work with queue, worker, payload, priority, and next run posture.

Schedule a safe retry with reason, attempt, and backoff.

Reserve a safe key so retries and replays cannot duplicate work.

Check whether a key exists before repeating a sensitive action.

Queue provider event replay with idempotency protection.

Prepare failed-job recovery with selected strategy.

Record preflight result before risky replays, retries, restores, or payments.

Capture operational failure into the error inbox.

Mark an error as acknowledged without deleting evidence.

Open a reliability incident with severity and owner.

Close an incident with summary and owner.

Claim work safely so duplicate workers do not collide.

Release a job lease after completion or rollback.

Hold failed work safely for review or recovery.

Record controlled future or recurring work.

Capture current reliability posture for review.

Prepare reliability report across jobs, retries, errors, incidents, and recovery.

Search jobs, retries, keys, replays, recoveries, errors, incidents, and reports.