Where are the resulting lessons saved?

Lessons can be persisted in several ways depending on your choice, including tag notes for specific contexts, CLAUDE.md for project rules, or Claude Code's session-wide auto-memory.

Can I perform a postmortem if I didn't record many logs?

Yes, if data is sparse, the skill will notify you and allow you to manually reconstruct the steps through dialogue to ensure the review still takes place.

Does it require the cc-memory plugin?

Yes, this skill is part of the cc-memory suite and relies on activity logs, topics, and decisions recorded during your work sessions.

What is the primary purpose of the postmortem skill?

It is designed to help you look back at completed activities step-by-step to understand what happened, why it happened, and how to improve future performance by saving those lessons.

How does the AI handle the review process?

The AI acts as a factual facilitator, presenting chronological data and context while inviting the user to provide the actual evaluation, ensuring judgment is user-led and accurate.

Activity Postmortem Analysis

Name: Activity Postmortem Analysis
Author: isizono

byisizono

0•

生産性とワークフロー

Facilitates structured, step-by-step reviews of completed activities to extract and persist actionable lessons learned.

The postmortem skill provides a systematic framework for reflecting on completed tasks within the cc-memory ecosystem. It guides users through a chronological breakdown of actions, gathering context from logs, decisions, and materials to facilitate a collaborative review where the agent presents facts and the user evaluates outcomes. This ensures that development experiences are not lost but are instead converted into permanent knowledge via Claude's memory, tag notes, or project-wide rules to prevent the repetition of past mistakes and optimize future workflows.

主な機能

01Interactive step-by-step review process with user-led evaluation

02Data-driven context gathering from related topics and decisions

030 GitHub stars

04Automated identification of success patterns and improvement areas

05Chronological activity reconstruction using cc-memory logs and data

06Multi-channel lesson persistence including tag notes, CLAUDE.md, and auto-memory

ユースケース

01Conducting structured retrospectives after completing complex feature developments

02Refining development workflows by codifying successful experimental patterns into permanent rules

03Analyzing critical bug fixes to document root causes and prevent regressions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add isizono/cc-memory postmortem

For use in Claude.ai and ChatGPT