← Back to Skills Marketplace
256
Downloads
0
Stars
15
Active Installs
1
Versions
Install in OpenClaw
/install self-improving-agent-3-0-0
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 appears to do what it says: provide reminders, capture errors/corrections to .learnings/, and help promote repeatable learnings. Before installing or enabling hooks globally, review the scripts (activator.sh, error-detector.sh, extract-skill.sh) yourself and consider: (1) prefer project-level installation over user-wide (~/.openclaw) if you want to limit injection to a single project, (2) enabling hooks will make the reminder/error detector run in sessions with the agent's permissions—they are lightweight and only emit text or write local files when you run extract-skill.sh, but confirm you trust the skill source, and (3) no credentials are requested, but be cautious about promoting any captured content to shared workspace files if it could include sensitive data. If you want additional assurance, run the scripts in a sandboxed/project directory first and inspect created files before copying them into your global OpenClaw folders.
Capability Analysis
Type: OpenClaw Skill
Name: self-improving-agent-3-0-0
Version: 1.0.0
The 'self-improving-agent' skill bundle is a legitimate development tool designed to facilitate continuous learning for AI agents by logging errors, corrections, and insights into markdown files. It includes utility scripts (e.g., `extract-skill.sh`, `error-detector.sh`) and OpenClaw hooks (`handler.js`) that inject reminders into the agent's context to encourage the capture of learnings. The scripts perform standard file system operations consistent with their stated purpose and include basic security measures, such as path traversal checks in the skill extraction script. No evidence of malicious intent, data exfiltration, or unauthorized execution was found.
Capability Assessment
Purpose & Capability
The name/description (self-improvement / log learnings) match the included files: a bootstrap hook, lightweight reminder and error-detection scripts, templates, and a skill-extraction helper. All artifacts are directly related to capturing and promoting learnings and integrating with OpenClaw/Claude-style hooks.
Instruction Scope
SKILL.md focuses on creating .learnings/ logs, copying hooks into OpenClaw (~/.openclaw/hooks), and enabling optional scripts (activator, error-detector). This is within scope, but installing at user/global level will cause the workspace files and hook behavior to be injected into every session — the user should be aware this promotes reminders across sessions. The error-detector reads CLAUDE_TOOL_OUTPUT (expected), and extract-skill.sh writes files locally when run (expected).
Install Mechanism
No automatic install or remote downloads are declared. The skill is instruction-only with local scripts and hooks. All scripts operate on local files and use standard shell utilities; no network fetches or obscure installers are present.
Credentials
The skill declares no required environment variables or credentials. The only environment usage is reading CLAUDE_TOOL_OUTPUT in the error-detector (documented and relevant). No secrets or unrelated service credentials are requested.
Persistence & Privilege
always:false (normal). The hook/installation recommendations encourage placing files under ~/.openclaw and enabling a hook; that gives the skill workspace-wide presence if the user opts in. Hooks run with the same permissions as the agent, so enabling them at the user/global level affects all sessions — this is expected but worth noting before global enablement.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install self-improving-agent-3-0-0 - After installation, invoke the skill by name or use
/self-improving-agent-3-0-0 - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of the self-improvement skill for OpenClaw and other agent platforms.
- Enables structured logging of learnings, errors, and feature requests to Markdown files for continuous improvement.
- Provides workflow for promoting important learnings to project memory files (`CLAUDE.md`, `AGENTS.md`, etc.).
- Includes detailed templates and quick reference guides for different types of entries (learnings, errors, feature requests).
- Supports OpenClaw-specific integration, workspace structure, and optional automatic reminder hook.
- Offers generic setup instructions for use in non-OpenClaw agent environments.
Metadata
Frequently Asked Questions
What is Self Improving Agent 3.0.0?
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 256 downloads so far.
How do I install Self Improving Agent 3.0.0?
Run "/install self-improving-agent-3-0-0" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Self Improving Agent 3.0.0 free?
Yes, Self Improving Agent 3.0.0 is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Self Improving Agent 3.0.0 support?
Self Improving Agent 3.0.0 is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Self Improving Agent 3.0.0?
It is built and maintained by 991200448 (@991200448); the current version is v1.0.0.
More Skills