self-improvement

Log errors, learnings, and feature requests to markdown files.

Updated Feb 18, 2026
One-click install
npx skills add https://github.com/ollieb89/openrepo --skill self-improvement-ollieb89
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: self-improvement
Source: https://github.com/ollieb89/openrepo/tree/main/skills/self-improving-agent
Command: npx skills add https://github.com/ollieb89/openrepo --skill self-improvement-ollieb89

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill helps AI agents and developers continuously improve by systematically capturing learnings, errors, and user corrections, preventing recurring mistakes and enhancing future performance.

Core Features & Use Cases

  • Error Logging: Records command failures, exceptions, and unexpected behaviors.
  • Learning Capture: Logs corrections, insights, and knowledge gaps from interactions.
  • Feature Request Tracking: Documents user-requested capabilities.
  • Promotion: Facilitates promoting valuable learnings to project memory (e.g., CLAUDE.md, AGENTS.md) or extracting them as new skills.
  • Use Case: After a complex debugging session where an obscure error was resolved, log the error, its cause, and the fix to .learnings/ERRORS.md. Later, if the pattern repeats, promote it to TOOLS.md or extract it as a new skill.

Quick Start

Log the recent command failure to the error log.

Frequently Asked Questions about self-improvement

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

FAQPage Schema
How do I log errors and learnings from AI agent interactions for continuous improvement?

To log errors and learnings for continuous improvement, you can systematically record command failures and user corrections into structured markdown files within a `.learnings/` directory. This prevents recurring mistakes and enhances future agent performance.

What is the best way to track feature requests and knowledge gaps discovered during debugging?

The best way to track feature requests and knowledge gaps is to document them during interactions and save the insights into structured markdown logs. This allows you to systematically address user-requested capabilities and bridge missing knowledge.

Can I promote resolved debugging insights to project memory files like CLAUDE.md?

Yes, you can promote resolved debugging insights to project memory files like CLAUDE.md or AGENTS.md. This mechanism transfers valuable learnings from local markdown logs into persistent project documentation for future reference.

Does capturing learnings in markdown files work for extracting reusable AI agent skills?

Capturing learnings in markdown files does work for extracting reusable AI agent skills. By organizing errors and corrections in a `.learnings/` directory, you can identify patterns and promote them into standalone skills for future tasks.

When should I extract a resolved error pattern as a new skill instead of just logging it?

You should extract a resolved error pattern as a new skill when the pattern repeats or proves broadly applicable. If a logged debugging insight is valuable, extracting it as a skill prevents the error from recurring across future interactions.