authoring-release-runbook — SKILL.md

Variant: standard · When to use: authoring the go-to-production operational runbook for a release — to a bar an unfamiliar on-call engineer can execute the deploy, confirm it, and roll back from the runbook alone.

Overview

This skill is the how-to of writing a strong release runbook — the operational procedure a deploying or on-call engineer follows to ship a system to production safely and verifiably. It is the broad operational go-live document: prerequisites and sign-offs, pre-deploy go/no-go checks, the ordered deploy procedure, post-deploy verification, the rollback procedure, and the on-call/escalation + monitoring pointers. This skill carries the producer's judgment — an SRE/operational research method and an executability/safety bar — not the section list. It assumes two collaborators: a release-runbook template tool that supplies the section structure, and a deep-research capability to ground the procedure in established practice. The producer is handed the upstream documents the plan determined inform this runbook (typically the architecture-doc + technical-design + test-plan) and must derive a real, executable procedure from them — never emit a generic skeleton. The bar to clear: an engineer who has never deployed this system can execute the release, confirm it succeeded, and roll back safely from the runbook alone, every step has an expected result and every forward change has a documented revert, no secret is inlined, and nothing is fabricated.

When to activate

• Authoring a production release/deployment/go-live runbook for a system or service.
• Documenting the manual + verification + rollback procedure around a deploy: prerequisites, go/no-go, ordered deploy steps, post-deploy smoke, rollback, escalation, and monitoring pointers.
• Filling a release-runbook template with researched, executable, upstream-derived content.

Do NOT activate when: