Postmortem

Lean personal Claude Code skills pack — my conventions, task flow, and quality lenses. Requires the superpowers plugin.

Install

npx -y skills add Endika/eskills --skill postmortem

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use when capturing an important bug or production failure — record the incident, root cause, fix, and lesson as an archivable markdown document.

SKILL.md

1.8 KB, as published. Nobody here has run it

postmortem

Overview

After an important bug or failure, capture what happened and why so the same class of problem can't quietly recur. Blameless: the target is the system and the gap, not who typed it. The output is one archivable markdown file.

When to use

A production failure, data issue, broken release, or a bug that cost real time.
NOT every small hiccup — a typo fixed in 30 seconds doesn't earn a postmortem.

Steps

Find the real root cause first. Use superpowers:systematic-debugging — don't document the symptom or the first plausible guess. The postmortem is only as good as the cause it names.
Fill the template from evidence, not memory: the actual error output, the commit/PR that fixed it, and the timeline if order mattered.
Make prevention concrete. "Be more careful" is not prevention. Name the specific thing that would have caught or blocked it: a test, a guard/validation, a CI gate, a CLAUDE.md note, or a durable memory.
Save it outside the repo: ../<repo>-notes/postmortems/YYYY-MM-DD-<slug>.md. Internal docs are never committed (see standards) — archive it, don't ship it.

Template

# Postmortem: <short title>

- **Date:** YYYY-MM-DD
- **Severity:** low | medium | high
- **Impact:** who/what was affected, for how long

## Incident

<one paragraph: the observable failure>

## Timeline (only if order mattered)

- HH:MM — …

## Root cause

<the underlying cause found via systematic-debugging — the "why", not the symptom>

## Fix

<what changed; commit / PR reference>

## Prevention

<the specific thing that stops this class of bug recurring>