Self Improvement

Logs mistakes, user corrections, and better approaches to markdown files so you can actually learn from sessions instead of repeating the same errors. Every time a command fails, the user says "no that's wrong," or you discover your knowledge is outdated, it goes into structured logs with IDs, priority, and suggested fixes. The OpenClaw integration is the star here: learnings get promoted to workspace files like SOUL.md or AGENTS.md that inject into every future session, and there are tools to share learnings across parallel agent sessions. It's maintenance overhead, but if you're running long projects or multiple agents, having a persistent memory layer beats starting from zero every time.

skills.sh installs: 204
GitHub stars: 674
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add sundial-org/awesome-openclaw-skills --skill self-improvement --agent claude-code

Installs into .claude/skills of the current project.

Files

Self-Improvement Skill

Log learnings and errors to markdown files for continuous improvement. Coding agents can later process these into fixes, and important learnings get promoted to project memory.

Quick Reference

Situation	Action
Command/operation fails	Log to `.learnings/ERRORS.md`
User corrects you	Log to `.learnings/LEARNINGS.md` with category `correction`
User wants missing feature	Log to `.learnings/FEATURE_REQUESTS.md`
API/external tool fails	Log to `.learnings/ERRORS.md` with integration details
Knowledge was outdated	Log to `.learnings/LEARNINGS.md` with category `knowledge_gap`
Found better approach	Log to `.learnings/LEARNINGS.md` with category `best_practice`
Similar to existing entry	Link with `See Also`, consider priority bump
Broadly applicable learning	Promote to `CLAUDE.md`, `AGENTS.md`, and/or `.github/copilot-instructions.md`
Workflow improvements	Promote to `AGENTS.md` (OpenClaw workspace)
Tool gotchas	Promote to `TOOLS.md` (OpenClaw workspace)
Behavioral patterns	Promote to `SOUL.md` (OpenClaw workspace)