A runbook exists
The service has written steps for its most likely failures, so the person on call does not start from zero.
HighChecked automatically
- Applies to:
- Every service
- Rule id:
- ownership.runbook-present
Why it matters
The person paged at 3 a.m. is often not the person who wrote the service. A runbook turns "figure it out" into "follow these steps", which shortens the outage and lowers the stress. Writing it is also a design review: a failure you can't describe a response to is a failure you haven't designed for.
How to apply it
- Cover the failures you expect first: a dependency is down, the queue is backing up, the database is full, a bad deploy needs rolling back.
- For each one give the symptom, how to confirm it, the action to take and how to tell it worked.
- Link every alert to its runbook entry.
- Review the runbook after each incident; an entry that was wrong during an incident is worse than none.
Example
"Order queue lag above 5 minutes. Check the consumer count on the orders dashboard. If consumers are healthy, scale the deployment to 6 replicas. If they are crash-looping, roll back the last release."
References
Checklists with this rule
Related agent skill: documentation-standards