skills.rest
Zhow01
Zhow01COMMUNITY
Explore all skills by this author →
GitHub

self-improvement

Log learnings, errors, and corrections from agent interactions into structured templates.

52|3|Updated Apr 3, 2026
Search Tags:#logging#self-improvement#hooks#learning#memory#prompting#openclaw
One-click install
npx skills add https://github.com/Zhow01/SkillAttack --skill self-improvement-zhow01
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: self-improvement
Source: https://github.com/Zhow01/SkillAttack/tree/main/data/hot100skills/001_pskoett_self-improving-agent
Command: npx skills add https://github.com/Zhow01/SkillAttack --skill self-improvement-zhow01

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 agent tasks. It provides templates and hooks to centralize guidance and promote important learnings to project memory.

Core Features & Use Cases

  • Log learnings to the .learnings directory, including corrections and insights
  • Record errors and feature requests for review and cross-session promotion
  • Promote recurring, high-value learnings to shared workspace memory (CLAUDE.md, AGENTS.md)
  • Integrate with OpenClaw hooks to prompt after prompts and tool usage

Quick Start

Enable the self-improvement hook and begin logging learnings and errors as they occur.

Frequently Asked Questions about self-improvement

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

FAQPage Schema
How do I capture and log agent learnings across multiple sessions?▼

To log agent learnings across sessions, this Skill identifies non-obvious issues and corrections from daily interactions, then stores them in a structured .learnings directory for future reference and cross-session promotion.

How do I record tool failures and user corrections for future reference?▼

Recording tool failures and user corrections is handled by enforcing structured templates and safe logging during daily agent interactions, capturing actionable guidance while preventing the exposure of secrets.

Can I integrate hook prompts to capture errors after tool usage?▼

Hook integration is supported through OpenClaw hooks, which automatically prompt for logging learnings and errors immediately after agent prompts and tool usage occur during active sessions.

How do I promote recurring patterns into project memory files?▼

Promoting recurring patterns into project memory involves elevating high-value learnings from the local log directory into shared workspace files like CLAUDE.md and AGENTS.md for continuous improvement.

Does this logging approach work without exposing secrets or sensitive data?▼

Safe logging without exposing secrets is enforced by the Skill's core design, which applies structured templates to capture actionable guidance while filtering sensitive information during the knowledge gap recording process.

Related Skills

Explore More Software Engineering Skills→
aaron-he-zhuaaron-he-zhu
142

memory-management

Manage two-layer memory with hot cache and cold storage for SEO projects.

Community
Basic
binfen1binfen1

self-improvement

Log learnings and corrections to markdown files for AI coding agents.

Community
Advanced
wewetv1987-cellwewetv1987-cell

dory-memory

Record exact user instructions in file-based workspace state across session resets.

Community
Advanced
KushtrvediKushtrvedi

self-improvement

Logs corrections and insights to YAML-formatted files in .learnings directory.

Community
Advanced