skills.rest
family3253
family3253COMMUNITY
Explore all skills by this author →
WebsiteGitHub

self-improving-agent

Log learnings, errors, and corrections to markdown files with YAML frontmatter.

1|Updated Mar 15, 2026
Search Tags:#self-improvement#hooks#memory#learnings#log#promotions#openclaw
One-click install
npx skills add https://github.com/family3253/skill --skill self-improving-agent-family3253
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: self-improving-agent
Source: https://github.com/family3253/skill/tree/main/skills/self-improving-agent
Command: npx skills add https://github.com/family3253/skill --skill self-improving-agent-family3253

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

The Self-Improvement Skill logs learnings, errors, and corrections to enable continuous improvement across sessions and tools.

Core Features & Use Cases

  • Log corrections to LEARNINGS.md with context and categories.
  • Capture errors to ERRORS.md and feature requests to FEATURE_REQUESTS.md for traceability.
  • Promote broadly useful learnings to CLAUDE.md, AGENTS.md, or copilot-instructions.md and optionally extract new skills.
  • Supports OpenClaw workspace integration and hook-based reminders to drive discipline.

Quick Start

Log a learning entry after completing a task or when an error occurs to begin your continuous-improvement cycle.

Frequently Asked Questions about self-improving-agent

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

FAQPage Schema
How do I persist AI session errors and corrections for continuous improvement?▼

To persist AI session errors and corrections, log entries into LEARNINGS.md and ERRORS.md files with context and categories to maintain a continuous-improvement record across sessions.

What is the best way to capture failed commands and knowledge gaps from AI sessions?▼

Capturing failed commands and knowledge gaps requires organizing in-session prompts and corrections into structured logs under a dedicated .learnings directory for later extraction.

How do I promote broadly useful learnings to project memory files like CLAUDE.md?▼

Promoting useful learnings to project memory files involves extracting validated entries from your logs and writing them into CLAUDE.md, AGENTS.md, or copilot-instructions.md.

Does the self-improving agent work with OpenClaw workspaces and hook-based reminders?▼

Yes, the self-improving agent supports OpenClaw workspace integration and hook-based reminders to drive logging discipline and ensure learnings are captured systematically.

What format do I need to structure learning logs for AI agents?▼

Structuring learning logs requires a YAML frontmatter with name and description fields, ensuring entries are organized and persist correctly within the .learnings directory.

Can I extract new skills from recurring corrections and feature requests?▼

Yes, you can extract new skills by promoting recurring corrections and feature requests from FEATURE_REQUESTS.md logs into standalone skill definitions for future use.

Related Skills

Explore More Software Engineering Skills→
benredmondbenredmond
3

compound

Parse the task file and append session learnings under future-agent-notes.

Community
Advanced
zai-orgzai-org
588

self-improvement

Capture AI session learnings, errors, and corrections into a .learnings repository.

Official
Advanced
handsomelong922handsomelong922

self-improvement

Log learnings and corrections to .learnings markdown files for future sessions.

Community
Advanced
martineseriosmartineserios
3

close

Analyze git logs and file changes to generate session summaries and detect documentation drift.

Community
Advanced