retro-create

Analyze skill invocation results and log post-run issues to retro.md.

3|1|Updated Feb 12, 2026

Search Tags:#automation #debugging #logging #self-improvement #retrospective #root-cause #skill-improvement

Ready to run

npx skills add https://github.com/edfenton/claude-skills --skill retro-create

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: retro-create
Source: https://github.com/edfenton/claude-skills/tree/main/shared/retro-create
Command: npx skills add https://github.com/edfenton/claude-skills --skill retro-create

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Post-invocation retrospective to identify and fix recurring issues from skill runs, improving reliability and learning over time.

Core Features & Use Cases

Log post-run issues to retro.md for audit and trend analysis.
Identify root causes (failed commands, user corrections, wrong assumptions, missing steps, outdated instructions) and apply durable fixes.
Update the skill definitions to prevent recurrence and streamline future invocations.

Quick Start

After a skill finishes running, review retro.md, implement fixes, and commit the updated SKILL.md to ensure the improvement is carried forward.

Frequently Asked Questions about retro-create

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

FAQPage Schema

How do I automate post-run debugging for skill invocations?▼

Automated post-run debugging analyzes the results of every skill invocation, logs issues to retro.md, and applies durable fixes to prevent recurrence. It detects failed commands and wrong assumptions to identify root causes.

What is a post-invocation retrospective for skill improvement?▼

A post-invocation retrospective is an automated analysis process that logs post-run issues, identifies root causes like missing steps or outdated instructions, and updates skill definitions to improve reliability and learning over time.

How do I log root-cause analysis data from failed commands?▼

You log root-cause analysis data by automatically recording failed commands, user corrections, and wrong assumptions into retro.md, which serves as an audit log for trend analysis and immediate fixes.

Can I automatically fix wrong assumptions and outdated instructions in my skills?▼

Yes, you can automatically fix wrong assumptions and outdated instructions by applying durable updates to skill definitions. If a fix alters behavior, it marks the issue as open with a resolution hint for manual validation.

When should I require manual validation for automated skill fixes?▼

Manual validation is required when an automated fix would alter skill behavior. The system marks the issue as status: open with a resolution hint, ensuring you review the change before committing the updated SKILL.md.

What's the best way to prevent recurring issues from skill runs?▼

The best way to prevent recurring issues is to automatically detect failed commands and missing steps, apply immediate fixes, and commit the updated SKILL.md so the improvement is carried forward into future invocations.