Can I analyze only specific parts of my history?

Yes, you can use subcommands like 'retrospective wins' or 'retrospective problems' to focus on specific dimensions of your workflow.

What data does this skill access?

It examines your local Claude Code session logs stored in your project directory, covering the last 90 days of interaction history.

How does the Retrospective skill help my workflow?

It analyzes your interaction history to identify repetitive tasks, success patterns, and friction points, suggesting specific ways to automate your dev environment.

Does it suggest new skills to build?

Yes, it specifically looks for 'Skill Opportunities'—recurring patterns in your requests that are prime candidates for custom slash commands or automation.

How does it handle technical analysis during the run?

It uses a 'one script per agent' principle to efficiently parse JSONL logs and extract metadata without overwhelming the user with individual bash approvals.

AI Workflow Retrospective

Name: AI Workflow Retrospective
Author: florianbuetow

byflorianbuetow

•

Productividad y Flujo de Trabajo

Analyzes developer-AI interaction logs to provide actionable insights for workflow optimization and automated skill creation.

The Retrospective skill brings agile principles to AI-assisted development by systematically reviewing your Claude Code session history from the last 90 days. It evaluates five key dimensions—success patterns, failure points, skill opportunities, workflow automation, and collaboration antipatterns—to identify exactly how your partnership with Claude can be improved. By examining real usage data rather than hypothetical best practices, it provides personalized recommendations for new subagents, hooks, and slash commands tailored to your specific coding style and project needs.

Características Principales

01Sentiment detection to identify user frustration or friction points

029 GitHub stars

03Multi-dimensional reporting on success and failure patterns

04Automated identification of repetitive tasks for skill creation

05Comprehensive session log analysis and metadata extraction

06Historical progress tracking across multiple retrospective reports

Casos de Uso

01Optimizing repetitive coding tasks into custom automated skills

02Generating an inventory of recent development sessions and outcomes

03Identifying and resolving collaboration friction points with the AI

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add florianbuetow/claude-code retrospective

For use in Claude.ai and ChatGPT

Características Principales

01Sentiment detection to identify user frustration or friction points

029 GitHub stars

03Multi-dimensional reporting on success and failure patterns

04Automated identification of repetitive tasks for skill creation

05Comprehensive session log analysis and metadata extraction

06Historical progress tracking across multiple retrospective reports

Casos de Uso

01Optimizing repetitive coding tasks into custom automated skills

02Generating an inventory of recent development sessions and outcomes

03Identifying and resolving collaboration friction points with the AI

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add florianbuetow/claude-code retrospective

For use in Claude.ai and ChatGPT