---
name: "runbook-skeleton"
description: "Turn a postmortem into an incident runbook with explicit verification gates. Use when the user needs: Get detection, diagnosis, mitigation, recovery and escalation sections."
metadata: {"version": "2026-10-17.1", "category": "productivity-ops", "publisher": "AIDiveForge"}
---

# Runbook Skeleton

Get detection, diagnosis, mitigation, recovery and escalation sections.

## Requirements

Provide the incident record, supported environment and authorized operational actions.

## Workflow

1. Extract the observed symptoms, affected systems, confirmed cause and effective recovery steps.
2. Separate diagnostic reads from state-changing mitigation; name the target environment for commands.
3. For each intervention state the trigger, expected effect, verification and rollback limit.
4. Add escalation points for uncertainty, failed recovery or actions outside the operator’s authorization.
5. Mark untested commands and missing details instead of inventing production-specific procedures.

## Worked example

Read `examples/input.txt` and `examples/expected.txt` before adapting the workflow.

## Acceptance checks

Use the confirmed rollback outcome while leaving absent command syntax and thresholds unresolved.

## Limits

A generated runbook is a draft until operators test it in the intended environment.

Worked example included. Model-specific performance and vendor compatibility have not been benchmarked.