self-improvement
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Claude ('No, that's wrong...', 'Actually...'), (3) User requests a capability that doesn't exist, (4) An external API or tool fails, (5) Claude realizes its knowledge is outdated or incorrect, (6) A better approach is discovered for a recurring task. Also review learnings before major tasks.
Why use this skill?
Enhance your AI agent's performance with the OpenClaw Self-Improvement skill. Automatically log errors, corrections, and best practices for smarter, more reliable workflows.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/fanghuaqi/self-improving-agent-v2What This Skill Does
The self-improvement skill for OpenClaw is a systematic meta-cognitive framework designed to turn every interaction, error, and correction into a structured learning asset. Instead of repeating the same mistakes or forgetting refined workflows, this skill enforces a discipline of logging outcomes into a dedicated .learnings workspace directory. By maintaining logs for LEARNINGS.md, ERRORS.md, and FEATURE_REQUESTS.md, the agent effectively creates a persistent feedback loop. This architecture allows OpenClaw to review its historical performance before tackling complex tasks, ensuring that previous corrections are applied proactively. It serves as an internal documentation engine that bridges the gap between ephemeral conversational context and long-term project knowledge.
Installation
Installation is streamlined via the ClawdHub ecosystem. To get started, execute clawdhub install self-improving-agent in your terminal. For manual configurations, clone the repository into ~/.openclaw/skills/self-improving-agent. Ensure that your workspace is initialized with the required directory structure by running mkdir -p ~/.openclaw/workspace/.learnings. This creates the necessary environment for the skill to begin writing and reading your project-specific wisdom.
Use Cases
This skill is indispensable for long-running development projects. It is triggered when an external API call fails, allowing the agent to document the error details for later debugging. It is equally powerful when the user provides direct feedback, such as correcting an incorrect function implementation. By capturing the correction immediately, the agent avoids re-introducing the same error in future sessions. Furthermore, it tracks feature requests that fall outside current capabilities, preventing the frustration of missing features from being lost in the noise of a long chat history.
Example Prompts
- "I've noticed you keep using the deprecated API method; please log this as a best_practice learning and update the current implementation."
- "That last command failed with a 403 error. Can you log this in the ERRORS.md file with the integration details so we can investigate it later?"
- "Review the current .learnings folder before we start this new refactor to ensure we don't repeat past mistakes."
Tips & Limitations
For maximum efficacy, treat your .learnings folder as a living codebase. Periodically "promote" your learnings. If a note in LEARNINGS.md becomes a standard procedure, move it to AGENTS.md or SOUL.md. Remember that the agent can only learn as much as you document; consistently prompting the agent to log failures is key to building a robust, self-correcting autonomous assistant.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-fanghuaqi-self-improving-agent-v2": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: file-write, file-read