retrospective

Analyze retrospective data to identify recurring mistakes and generate action items.

4|Updated Sep 25, 2025
One-click install
npx skills add https://github.com/cowwoc/styler --skill retrospective
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: retrospective
Source: https://github.com/cowwoc/styler/tree/main/.claude/skills/retrospective
Command: npx skills add https://github.com/cowwoc/styler --skill retrospective

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Analyze accumulated mistakes, identify recurring patterns, check action item effectiveness, and generate new action items for systemic fixes.

Core Features & Use Cases

  • Identify patterns across mistakes and categorize by type.
  • Assess action item effectiveness and propose escalations when needed.
  • Produce actionable recommendations to prevent recurrence.

Quick Start

/workspace/main/.claude/scripts/retrospective.sh

Frequently Asked Questions about retrospective

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

FAQPage Schema
How do I identify recurring mistakes and patterns in my work?

Retrospective analysis identifies recurring problems by categorizing mistakes, extracting patterns and keywords, then generating actionable fixes. The Skill processes accumulated mistake logs to surface systemic issues that need addressing.

Can I automatically check whether my action items are actually working?

Yes. The Skill evaluates action-item effectiveness against accumulated mistake data, proposes escalations when fixes fail, and generates updated recommendations based on whether problems recur.

How do I get started with mistake analysis and retrospectives?

Run `/workspace/main/.claude/scripts/retrospective.sh` to invoke the Skill manually, or let it run automatically via the SessionStart trigger. It reads from `.claude/retrospectives/mistakes.json` and `.claude/retrospectives/retrospectives.json` to produce updated retrospectives and action items.

What's the difference between categorizing mistakes by type versus extracting patterns?

Categorizing by type organizes mistakes into structured groups; extracting patterns identifies recurring keywords and themes across categories. Both work together to reveal systemic root causes rather than isolated incidents.

When should I escalate a mistake instead of creating another action item?

Escalate when an action item fails to prevent a mistake from recurring. The Skill generates escalation entries automatically when fixes don't resolve underlying problems, signaling that a deeper systemic fix is needed.

Do I need to manually format mistake data, or does the Skill handle JSON conversion?

The Skill operates directly on `.claude/retrospectives/mistakes.json` and `.claude/retrospectives/retrospectives.json` without requiring manual preprocessing. It exports results back to the same logs in structured JSON format.