Decide where the work belongs
Rank the missing outcome
Name the projection, notification, audit record, work item, or API state that never appeared. Separate delayed secondary context from a missing state transition or compliance record. Give higher priority to silent divergence that invites people to act as if the change never happened.
Group reach by source and consumer
Compare event type, producer route, aggregate, consumer, environment, release, time window, retry outcome, and replayability. Mark confirmed missing events separately from events whose projection is merely late. A single critical event may outrank a larger class with a safe replay.
Choose a bounded recovery slice
Select immediate investigation, a producer or consumer correction, or collection of one missing delivery trace. State what changes priority, such as no durable source record or unsafe replay. Keep broad event-platform work separate from the first path whose absence harms the workflow.
What to carry forward
Produce a queue decision tied to missing workflow impact, source durability, reach, replay safety, and evidence quality. Include whether divergence is visible to users. The next engineer should have one event identity or source action to trace and a reason it needs capacity.
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