--- name: runbook-auditor description: "Runbook auditor: read runbook files, extract verifiable claims, check each against current codebase. LLM judges correctness and completeness. Read-only. Trigger: /runbook-audit" trigger: /runbook-audit --- # /runbook-audit Runbooks are the first casualty of shipping. This skill audits them for correctness. ## What this is for - Commands that reference nonexistent services or wrong names - Dashboards that redirect or no longer exist - Recovery procedures missing critical steps - **Read-only skill.** No runbook modification. ## PROTECTION RULE - never ~/.claude/ Read-only skill. Guard required if write mode added later. ## What You Must Do When Invoked During analysis, assign a confidence level to each finding: proven (confirmed by evidence), likely (strong signal, needs review), or suspected (weak signal). If `/runbook-audit -help` or `/runbook-audit -h` (without further arguments) is invoked: output the `## Usage` section unchanged and stop. Otherwise follow these steps in order, skipping none. ### Step 1 - Help check If invoked with `-help` or `-h`, output the `## Usage` section unchanged and stop. ### Step 2 - Clarify target Clarify `-ProjectDir`. Get confirmation. ### Step 3 - Scan ```powershell & "/scripts/runbook-read.ps1" -ProjectDir "" ``` ### Step 4 - Analysis Read each runbook claim: - Do referenced commands match current docker-compose/CI/infra config? - Do dashboard URLs exist in the codebase? - Are recovery procedures complete and in order? - Are contacts and escalation paths valid? ### Step 5 - Write report File `runbook-quality-report.md` in current working directory: 1. **Summary** - freshness score, total claims, valid, stale, invalid. 2. **Per-runbook analysis** - file, score, findings. 3. **Recommendations** - specific fixes for stale/invalid claims. 4. **Open questions**. ### Step 6 - Summarize State report path, highlight most out-of-date runbook. ## Usage ``` /runbook-audit # interactive /runbook-audit # scan project /runbook-audit -help ```