What problem does it solve?
AI coding agents and development teams lose valuable lessons from unexpected errors, user corrections, and outdated knowledge when these insights are not systematically captured and shared, leading to repeated mistakes and wasted effort across projects and sessions.
Core Features & Use Cases
- Structured Logging: Automatically log errors, user corrections, knowledge gaps, and feature requests to standardized .learnings/ files with consistent formatting, priority levels, and area tags for easy filtering and review.
- Knowledge Promotion: Elevate high-value, broadly applicable learnings to project-wide memory files like CLAUDE.md, AGENTS.md, and .github/copilot-instructions.md to prevent recurring issues for all team members and AI agents.
- Recurring Pattern Detection: Identify repeated errors or corrections, link related entries, and trigger skill extraction for solutions that are useful across multiple codebases.
- Use Case: If your team repeatedly encounters the same CI/CD pipeline timeout error, this skill logs each occurrence, links the related entries, and helps you promote the fix to your team's AGENTS.md file so all agents avoid the issue in the future.
Quick Start
Ask your AI coding assistant to log the recent command failure you encountered when running the test suite to your project's .learnings/ERRORS.md file so you can reference the fix later.
