What problem does it solve?
This skill provides a consistent, low-friction way to capture operational learnings, user corrections, command failures, and feature requests during AI-assisted development sessions so those insights are preserved, reviewed, and promoted into project memory or reusable skills.
Core Features & Use Cases
- Structured Logging: Standardized markdown templates for LEARNINGS, ERRORS, and FEATURE_REQUESTS to keep entries uniform and searchable.
- Hook-Based Detection: Lightweight hook scripts that remind agents to evaluate sessions and detect command/tool errors automatically.
- Promotion Workflows: Guidance and scripts to promote recurring or high-value learnings into CLAUDE.md, AGENTS.md, or extracted skills for broader reuse.
- Integration Scenarios: Use in interactive agent sessions, CI-assisted runs, or developer workspaces to capture post-task discoveries and recurring gotchas.
Quick Start
After finishing a task, log any error, correction, or useful insight to .learnings/LEARNINGS.md or .learnings/ERRORS.md using the provided template.