← Back to Skills Marketplace
341
Downloads
0
Stars
6
Active Installs
1
Versions
Install in OpenClaw
/install self-improving-agent-v2
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: remind agents to log learnings, detect simple command errors, and scaffold extracted skills. Before installing or enabling hooks: (1) review the scripts (activator.sh, error-detector.sh, extract-skill.sh) — extract-skill.sh creates files/directories in the current workspace; (2) if you enable the hook by copying files to ~/.openclaw/hooks and configuring agent settings, those scripts will run with the agent's permissions, so enable only if you trust their content; (3) the error detector reads CLAUDE_TOOL_OUTPUT (agent-provided output) — no external exfiltration is present, but confirm you don't wire hooks to remote endpoints; (4) prefer installing from a repository you trust or inspect the repo contents locally. The only minor doc inconsistency: one doc claims scripts only output text but extract-skill.sh does write files — treat that as intentional functionality and review the script before running.
Capability Analysis
Type: OpenClaw Skill
Name: self-improving-agent-v2
Version: 1.0.0
The skill bundle implements a 'self-improvement' framework designed to help AI agents log errors, user corrections, and new patterns into local markdown files (.learnings/) and workspace memory (SOUL.md, AGENTS.md). The included bash scripts (activator.sh, error-detector.sh, extract-skill.sh) and OpenClaw hooks (handler.js/ts) are standard utility scripts for context injection and file scaffolding; they contain appropriate path validation to prevent traversal and show no signs of data exfiltration, unauthorized network access, or malicious intent.
Capability Assessment
Purpose & Capability
Name/description (capture learnings/errors/promotions) align with the included files: logging templates, hook handlers, activator/error-detector scripts, and a skill-extraction helper. Required env/config/binaries are minimal or absent, matching the stated purpose.
Instruction Scope
SKILL.md and hook docs stay within the declared purpose (create .learnings/, inject reminders, run small hook scripts). One doc claim in references/hooks-setup.md states "scripts only output text; they don't modify files or run commands", but scripts include extract-skill.sh which creates directories/files (intended behavior). Also error-detector.sh reads the CLAUDE_TOOL_OUTPUT env var (agent-provided) — expected but worth noting. Overall scope is appropriate; review extract-skill.sh if you don't want the skill to create files.
Install Mechanism
No install spec in registry (instruction-only). Manual install suggestions use git clone from a GitHub repo (traceable). No downloads from untrusted hosts or archive extraction are present.
Credentials
The skill requests no credentials or sensitive env vars. Scripts use CLAUDE_TOOL_OUTPUT and event/context objects provided by the agent platform — appropriate for hook behavior. No unrelated secrets or config paths are requested.
Persistence & Privilege
always:false and user-invocable default. The skill can add a hook to OpenClaw if you copy/enable it (user action). The extract-skill.sh creates files under a relative ./skills path (script protects against absolute paths and '..') — this is normal for a scaffold helper but you should run it consciously.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install self-improving-agent-v2 - After installation, invoke the skill by name or use
/self-improving-agent-v2 - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of the self-improving-agent skill for continuous learning and error logging.
- Provides clear workflows for logging errors, user corrections, feature requests, and knowledge gaps to markdown files.
- Defines standardized formats and fields for learnings, errors, and feature request entries.
- Guides on promoting important insights to project memory files (SOUL.md, AGENTS.md, TOOLS.md, etc).
- Integrates with OpenClaw for automated workspace setup and inter-session learning sharing.
- Includes setup instructions for both OpenClaw and generic agents.
Metadata
Frequently Asked Questions
What is Self Improving Agent V2?
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 341 downloads so far.
How do I install Self Improving Agent V2?
Run "/install self-improving-agent-v2" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Self Improving Agent V2 free?
Yes, Self Improving Agent V2 is completely free (open-source). You can download, install and use it at no cost.
Which platforms does Self Improving Agent V2 support?
Self Improving Agent V2 is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Self Improving Agent V2?
It is built and maintained by hacker (@fanghuaqi); the current version is v1.0.0.
More Skills