← Back to Skills Marketplace
Self Improvement
by
zhihongyee
· GitHub ↗
· v0.1.0
· MIT-0
104
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install self-improvement-2
Description
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
Usage Guidance
This skill is coherent and appears to do what it says: log learnings and errors to markdown files. Before installing, consider these precautions: (1) Decide whether .learnings/ should be in source control — add it to .gitignore if you do not want logs in your repo history. (2) Ensure the agent or CI does not log secrets, PII, or credentials into these files; add redaction or sanitization steps to any automated logging. (3) Be cautious about the 'promote to project memory' guidance because that writes to shared files (CLAUDE.md, AGENTS.md, etc.); require human review before promotion. (4) The SKILL.md shows an npx install command — only run it if you trust the upstream package name/owner. (5) If you want tighter control, restrict which contexts or users can invoke this skill, and add obvious markers for sensitive entries so they are not published automatically.
Capability Analysis
Type: OpenClaw Skill
Name: self-improvement-2
Version: 0.1.0
The skill facilitates a self-improvement loop by logging errors and learnings to local files and promoting them to project-level documentation. It is classified as suspicious due to instructions that direct the agent to modify sensitive configuration files (e.g., `.claude/settings.json`) and execute unverified shell scripts (`scripts/activator.sh`, `scripts/error-detector.sh`) via automated hooks. These high-risk capabilities, while serving the stated purpose, create a significant attack surface for persistent prompt injection or unauthorized code execution if the learning process is manipulated.
Capability Assessment
Purpose & Capability
The name/description (capture learnings/errors and promote to project memory) matches the instructions: create a .learnings/ folder and append structured markdown entries. It does not request unrelated credentials or binaries.
Instruction Scope
Instructions are narrowly focused on creating and appending to files under .learnings/ and promoting entries to project files (CLAUDE.md, AGENTS.md, .github/copilot-instructions.md, etc.). This is within purpose, but promotion steps imply writing to shared repository files and possibly exposing logged content to contributors — review/sanitization guidance would be prudent.
Install Mechanism
No registry install spec is embedded in the skill bundle (instruction-only). SKILL.md includes an example npx install command for convenience; that is informational and not enforced by the skill metadata. No downloads/executables are installed by the skill itself.
Credentials
The skill requests no environment variables, no credentials, and no config paths. There are no opaque secret requirements inconsistent with its stated purpose.
Persistence & Privilege
always:false (default) and model invocation is allowed (normal). The skill writes files inside the project workspace (.learnings and promoted docs) which is consistent with its goal; it does not request system-wide persistence or modify other skills.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install self-improvement-2 - After installation, invoke the skill by name or use
/self-improvement-2 - Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.1.0
Initial release enables structured learning and error logging for continuous self-improvement.
- Introduces markdown-based logging for learnings, errors, and feature requests, supporting CI/headless workflows.
- Provides detailed templates and metadata for consistent capture of incidents and improvements.
- Outlines promotion process for broadly applicable learnings to persistent project knowledge files.
- Includes quick-reference tables for when and how to log or promote different types of information.
- Defines mechanisms for recurring pattern detection and guidance for resolving or promoting insights.
Metadata
Frequently Asked Questions
What is Self Improvement?
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau... It is an AI Agent Skill for Claude Code / OpenClaw, with 104 downloads so far.
How do I install Self Improvement?
Run "/install self-improvement-2" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Self Improvement free?
Yes, Self Improvement is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Self Improvement support?
Self Improvement is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Self Improvement?
It is built and maintained by zhihongyee (@zhihongyee); the current version is v0.1.0.
More Skills