Decide where the work belongs
Map pending consequence
For each endpoint, record status, failure reason, pending identities, workflow state, event age, receiver capacity, and known effects. Separate a delayed display update from an access or reconciliation transition that may require careful ordering.
Compare recovery paths
Assess repairing configuration, re-enabling without replay, sending a bounded replay, reconciling known effects, or keeping the endpoint paused. Record owner, capacity check, deduplication evidence, and the condition that makes each path unsafe.
Set recovery order
Queue endpoints with durable state at risk and uncertain completed effects first, then isolated delivery backlog and history cleanup. Add a fixture, review time, and escalation trigger. Keep old pending events provisional when identity or effect state is incomplete.
What to carry forward
Return a recovery queue with consequence, backlog scope, containment, owner, and replay proof. Escalate unknown completed effects before bulk action. Keep rank provisional when disable reason, receiver capacity, or pending event history remains unobserved. Include receiver capacity.
Technical background: Zendesk developer documentation.
Keep the decision with the work.
Use a Work Item in Aglet to record the problem, the evidence you have, and the next decision. Add an owner and priority, then keep updates in the discussion so the next person can follow the reasoning.
Create an account See the product workflow