improve-skill

Analyze Claude Code session transcripts to generate structured skill improvement prompts.

13|1|Updated Jan 9, 2026
One-click install
npx skills add https://github.com/mauromedda/agent-toolkit --skill improve-skill-mauromedda
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: improve-skill
Source: https://github.com/mauromedda/agent-toolkit/tree/main/skills/improve-skill
Command: npx skills add https://github.com/mauromedda/agent-toolkit --skill improve-skill-mauromedda

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill helps you analyze Claude Code session transcripts to identify gaps in existing skills, capture valuable insights, and provide structured prompts to improve or create skills. It enables systematic reflection on past sessions to enhance documentation and workflows.

Core Features & Use Cases

  • Extracts and aggregates session transcripts for offline review.
  • Analyzes sessions to surface missing information, misalignment, and opportunities to improve skills.
  • Generates structured prompts to improve an existing skill or to create a new skill from a successful session.
  • Guides documentation updates with concrete recommendations and examples.

Quick Start

Run the provided scripts to extract a session transcript and generate an analysis or new-skill prompt:

  • uv run ~/.claude/skills/improve-skill/scripts/extract_session.py --latest --output /tmp/session.md
  • uv run ~/.claude/skills/improve-skill/scripts/analyze_session.py --transcript /tmp/session.md --skill ~/.claude/skills/your-skill/SKILL.md
  • uv run ~/.claude/skills/improve-skill/scripts/analyze_session.py --transcript /tmp/session.md --create-skill

Frequently Asked Questions about improve-skill

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

FAQPage Schema
How do I analyze Claude Code session transcripts to improve existing skills?

To analyze Claude Code session transcripts, run the extract_session.py script to output a markdown file, then run analyze_session.py with the transcript and target SKILL.md file to surface gaps and generate structured improvement prompts.

Can I generate a new skill template directly from a successful debugging session?

Yes, you can generate a new skill template from a debugging session by running the analyze_session.py script with the --create-skill flag against your extracted session transcript to produce structured creation prompts.

What is the best way to find missing information in my Claude Code automation workflows?

The best way to find missing information in automation workflows is to analyze session transcripts, which surface misalignment and documentation gaps by systematically reflecting on past Claude Code sessions.

Do I need Python to extract and review Claude Code session transcripts offline?

Yes, you need Python to run the provided scripts via the uv run command, which extracts the session transcript markdown and performs the offline analysis to identify skill improvement opportunities.

When should I use session transcript analysis instead of manually updating skill documentation?

You should use session transcript analysis after debugging or automation sessions when you need to systematically identify misalignment, capture insights, and generate concrete documentation updates rather than relying on manual review.

Why does prompt engineering matter for creating new skills from session learnings?

Prompt engineering matters because it transforms session learnings into structured prompts that guide documentation updates and new skill creation, ensuring that extracted insights translate into sharper, more effective skill definitions.