post-mortem · thananon — Claude Code skill: bug fix write-up with root cause
Skill writes an engineering post-mortem after debugging: root cause with code identifiers, symptom, mechanism, the fix, and why it slipped through. Use after validation, before closing the ticket.
npx skills add thananon/9arm-skills --skill post-mortem -a claude-codeWhat post-mortem does
post-mortem generates a complete engineering record of a bug fix for the team. It captures symptom, precise root cause with function names and file paths, how the cause leads to the observed failure, and the fix itself. The skill requires four things: reproducible repro, identified root cause (not a hypothesis), landed fix, and validation that the original failure now passes.
The report follows a fixed structure: summary, symptom, root cause, cause-to-symptom chain, fix, debugging path, why it slipped through, and validation. Each section is concrete — function names, code paths, commit SHAs. The skill refuses to draft if any of the four criteria are missing.
Tone is engineer-to-engineer, no softening. Blameless post-mortem: describes the CI gap, workload gap, or review gap, not the person. For leadership reframing, there's a separate skill that takes this record and restructures it for management.
What it helps with
- Write an engineering post-mortem with root cause, code identifiers, and commit links
- Document how the bug reached production and why tests missed it
- Leave a debugging breadcrumb trail for the next engineer
- Connect symptom to mechanism in the code
- Close a bug ticket with full fix history and validation steps
Example prompts for the agent
- Write the post-mortem for this bug with root cause and mechanism
- Document this fix: what broke, why, how we fixed it, how we validated
- Compose an RCA with the debugging path and why it slipped through
How to install post-mortem
1Skills CLI (Vercel)
npx skills add thananon/9arm-skills --skill post-mortem -a claude-code2Manually: put the files in place
~/.claude/skills/post-mortem/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/post-mortem/
1Skills CLI (Vercel)
npx skills add thananon/9arm-skills --skill post-mortem -a codex2Manually: put the files in place
.agents/skills/post-mortem/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add thananon/9arm-skills --skill post-mortem -a cursor2Manually: put the files in place
.cursor/skills/post-mortem/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install thananon/9arm-skills post-mortem2Manually: put the files in place
.github/skills/post-mortem/SKILL.md1Skills CLI (Vercel)
npx skills add thananon/9arm-skills --skill post-mortem -a gemini-cli2Platform CLI
gemini skills install https://github.com/thananon/9arm-skills --path skills/engineering/post-mortemManually: put the files in place
.devin/skills/post-mortem/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add thananon/9arm-skills --skill post-mortem -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
From the same set3
The whole set →Other versions of post-mortem
Similar
Questions about post-mortem
How do I install post-mortem in Claude Code?
Run npx skills add thananon/9arm-skills --skill post-mortem -a claude-code, then find the skill file at ~/.claude/skills/post-mortem/SKILL.md.
What does post-mortem do?
Writes a structured engineering post-mortem after you've debugged and fixed a bug. It includes symptom, root cause with code, cause-to-symptom chain, the fix, debugging path, analysis of why it slipped through, and validation.
When should I use post-mortem, and when should I skip it?
Use it after you have a reproducible repro, known root cause, landed fix, and validation. Skip it if the bug is not yet fixed, the fix is not validated, it's a production incident with paging history (that needs a separate incident report), or the fix is trivial (typo or one-liner — PR description is enough).
