Handoff Runbook

A handoff runbook is organised by symptom rather than by architecture: “documents stuck in queue” → check X, then Y, then Z, and if none of those, call this person. It is written for someone who has never seen the code, at 3am, under pressure — and every incident you handle should add an entry.

Worked example: the test of whether a runbook is real is not whether it exists but whether the customer’s engineer has led an incident using it while you watched. Gotcha: architecture diagrams are not runbooks; they explain how the system is built, which is not what a person diagnosing a stuck queue needs.