skills.rest
meain
meainCOMMUNITY
Explore all skills by this author →
Website𝕏@meain_GitHub

retrospect

Review conversations to derive lessons and propose CLAUDE.md guideline updates.

284|10|Updated Dec 6, 2015
Search Tags:#retrospective#ai improvement#lesson learned#claudem.md#conversation review#guideline generation
One-click install
npx skills add https://github.com/meain/dotfiles --skill retrospect-meain
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: retrospect
Source: https://github.com/meain/dotfiles/tree/main/claude/.claude/skills/retrospect
Command: npx skills add https://github.com/meain/dotfiles --skill retrospect-meain

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps improve AI performance by systematically reviewing past interactions, identifying errors and inefficiencies, and updating guidelines to prevent recurrence.

Core Features & Use Cases

  • Conversation Analysis: Scans chat history for agent mistakes, user corrections, and repeated errors.
  • Guideline Generation: Derives concise, actionable lessons from identified issues.
  • CLAUDE.md Updates: Proposes and applies new rules to a central knowledge base (CLAUDE.md) to prevent future mistakes.
  • Use Case: After a complex coding session where the AI made several small errors, this Skill can review the conversation, identify the patterns of mistakes, and suggest new rules to add to its core instructions to avoid similar issues in the future.

Quick Start

Review the last conversation and update CLAUDE.md with any lessons learned.

Frequently Asked Questions about retrospect

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I prevent AI from repeating the same coding mistakes across different conversations?▼

To prevent AI from repeating coding mistakes, you can review past conversations to identify errors and generate new guidelines. This skill analyzes chat history and proposes actionable rules to add to CLAUDE.md, ensuring the agent avoids similar issues in future sessions.

What is the best way to automatically update CLAUDE.md based on conversation corrections?▼

The best way to update CLAUDE.md automatically is by analyzing user corrections and agent inefficiencies within the current chat. The skill derives concise lessons from these interactions and directly applies them as new rules to your core knowledge base.

How does conversation review work for generating AI lesson learned guidelines?▼

Conversation review for generating guidelines works by scanning the chat history for agent mistakes and repeated errors. It then compares these actions against existing CLAUDE.md rules to derive specific, general, and brief lessons for future prevention.

Can I use conversation review to suggest new scripts for recurring workflows?▼

Yes, you can use conversation review to suggest new scripts for recurring workflows. Beyond updating CLAUDE.md, the skill identifies repeated patterns in the chat and proposes new skills or scripts to automate those workflows.

Do I need any external dependencies to run a retrospective on my AI chat history?▼

You do not need any external dependencies to run a retrospective on your AI chat history. The skill operates independently to review conversations, identify agent mistakes, and propose new guidelines without requiring additional components.

Related Skills

Explore More Software Engineering Skills→
dandudzidandudzi
2

review-claudemd

Analyze JSONL conversation logs to suggest CLAUDE.md file improvements.

Community
Advanced
daffy0208daffy0208
34

Manifest Generator

Generate YAML capability manifests from skill and MCP descriptions.

Community
Intermediate
ToanPV90ToanPV90

review-claudemd

Analyze recent JSONL conversation history to identify CLAUDE.md instruction improvements.

Community
Intermediate
Eric-Pacheco95Eric-Pacheco95

synthesize-signals

Synthesize project lessons into themes and candidate steering rules.

Community
Advanced