lucas-runbook — Claude Code skill: what it does, how to install
lucas-runbook forces the agent to troubleshoot Kubernetes failures using approved Markdown runbooks. Ideal for pod crash investigations and container restarts.
MIT 1 installs ★ 361 updated 2026-06-08
npx skills add a2wio/lucas --skill lucas-runbook -a claude-codeWhat lucas-runbook does
This skill locks the agent into runbook-driven mode for Kubernetes incidents. Whenever pod issues like CrashLoopBackOff, OOMKilled, or ImagePullBackOff appear, the agent searches for Markdown procedures in /runbooks using file patterns or fetches remote documentation with WebFetch.
Improvisation is banned. The assistant must cite the exact runbook path, run only authorized diagnostic commands, and apply specified fixes. If no document matches the issue or a runbook demands escalation, the agent halts and asks how to proceed.
What it helps with
- Investigate
OOMKilledpod errors using verified runbooks - Search local
/runbooksdirectory for matchingCrashLoopBackOffguides - Block unapproved remediation commands on Kubernetes clusters
- Fetch remote incident resolution procedures via
WebFetch
Example prompts for the agent
- A pod failed with OOMKilled, find the runbook and fix it according to procedure
- Investigate why the backend pod is in CrashLoopBackOff using /runbooks
- The deployment shows ImagePullBackOff, look up the remediation steps
How to install lucas-runbook
1Skills CLI (Vercel)
npx skills add a2wio/lucas --skill lucas-runbook -a claude-code2Manually: put the files in place
~/.claude/skills/lucas-runbook/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/lucas-runbook/
1Skills CLI (Vercel)
npx skills add a2wio/lucas --skill lucas-runbook -a codex2Manually: put the files in place
.agents/skills/lucas-runbook/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add a2wio/lucas --skill lucas-runbook -a cursor2Manually: put the files in place
.cursor/skills/lucas-runbook/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install a2wio/lucas src/agent/.claude/skills/lucas-runbook2Manually: put the files in place
.github/skills/lucas-runbook/SKILL.md1Skills CLI (Vercel)
npx skills add a2wio/lucas --skill lucas-runbook -a gemini-cli2Platform CLI
gemini skills install https://github.com/a2wio/lucas --path src/agent/.claude/skills/lucas-runbookManually: put the files in place
.devin/skills/lucas-runbook/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add a2wio/lucas --skill lucas-runbook -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about lucas-runbook
How to install lucas-runbook in Claude Code?
Run npx skills add a2wio/lucas --skill lucas-runbook -a claude-code. The skill writes to ~/.claude/skills/lucas-runbook/SKILL.md.
What does lucas-runbook do?
It forces your coding assistant to locate and follow official Markdown runbooks before applying any fixes to a Kubernetes environment.
What happens when no matching runbook exists?
The assistant will not attempt ad-hoc repairs. It reports observed errors and prompts you directly for instructions.
Source text
SKILL.md · MIT
--- name: lucas-runbook description: Lucas runbook-driven troubleshooting for Kubernetes issues. Use when investigating pod errors, crashes, restarts, or any Kubernetes problem. Automatically search runbooks before taking action. allowed-tools: Read, Glob, Grep, WebFetch --- # Lucas Runbook-Driven Troubleshooting You are operating in **runbook-driven mode**. Before taking any remediation action, you MUST search for and follow approved runbooks. ## Runbook Sources Search these locations for runbooks: ### Local Runbooks - Path: `/runbooks` - Search using: `Glob pattern="**/*.md" path="/runbooks"` - Read matching files to find relevant procedures ### External Documentation (if configured) - Check environment or config for external URLs - Use WebFetch to retrieve external runbook content ## Procedure When you identify a Kubernetes issue: 1. **Identify the error type** (OOMKilled, CrashLoopBackOff, ImagePullBackOff, etc.) 2. **Search for matching runbook**: ``` Glob pattern="**/*oom*.md" path="/runbooks" Glob pattern="**/*crash*.md" path="/runbooks" Glob pattern="**/*image*.md" path="/runbooks" ``` 3. **Read the runbook** if found 4. **Follow the runbook EXACTLY**: - Use only the approved diagnostic commands - Apply only the approved fixes - Escalate when the runbook says to escalate 5. **If NO runbook found**: - Do NOT attempt fixes - Report what you observed - Ask: "No runbook found for this issue. How should I proceed?" ## Important Rules - ALWAYS cite which runbook you're following - NEVER improvise fixes outside of runbooks - If a runbook says "escalate", do NOT attempt the fix yourself - Document what you did and the outcome ## Example When you see OOMKilled: 1. Search: `Glob pattern="**/*oom*.md" path="/runbooks"` 2. Find: `oom-killed.md` 3. Read: The runbook says to increase memory limits 4. Say: "Following runbook `oom-killed.md`: Increasing memory limit from 256Mi to 512Mi" 5. Execute the approved fix 6. Verify and report outcome
© a2wio · MIT · a2wio/lucas