A meta-skill designed to identify recurring patterns, user corrections, and repeated tool sequences to codify them into permanent, reusable AI capabilities.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install skill-learner
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install skill-learner using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Skill Learner?
Skill Learner is a strategic utility for developers using Openclaw Skills to evolve their AI agents from task-doers into pattern-recognizers. It focuses on meta-cognition by analyzing session history to detect when a user provides the same instruction multiple times or when a specific sequence of tool calls consistently leads to success or failure.
By formalizing these observations, Skill Learner ensures that the AI doesn't just complete a task, but learns the underlying procedural logic. This reduces technical debt and cognitive load for the user, making Openclaw Skills more autonomous and reliable over time.
Skill Learner Use Cases
- Standardizing project-specific workflows that require the same manual setup instructions every time.
- Reducing frequent user corrections by codifying those corrections into a permanent operating rule.
- Converting complex, multi-tool sequences into a streamlined, documented skill pattern.
- Identifying common failure points in existing Openclaw Skills to recommend targeted improvements.
How Skill Learner Works
- Pattern Detection: The agent monitors the environment for repeated signals, such as similar task requests or identical user feedback.
- Strategy Selection: The system determines if the pattern warrants a new standalone skill, an update to an existing one, or a simple local workspace rule.
- Extraction: The core logic, triggers, and necessary reference scripts are stripped from the noise of the conversation.
- Structural Recommendation: The agent defines the leanest possible implementation, ranging from a basic SKILL.md file to a full script-backed module.
- Reporting: A clear summary is generated detailing the observed pattern and the recommended technical path forward.
Skill Learner Setup
To integrate Skill Learner into your agent directory, ensure the skill definition is placed in your active skills path:
# Navigate to your skills directory
cd path/to/your/agent/skills
# Create the skill definition
touch skill-learner.md
Add the Skill Learner configuration to your system prompt to allow the agent to invoke this logic whenever it detects repetitive manual overhead.
Skill Learner Data Schema & Taxonomy
The Skill Learner organizes its findings into a structured report to ensure consistency across Openclaw Skills implementations:
| Component | Description |
|---|---|
| Pattern Observed | Details the specific repetition or recurring user correction. |
| Value Proposition | Explains the specific waste, error, or latency this skill prevents. |
| Recommended Action | Categorizes the output as a New Skill, Improved Skill, or Local Rule. |
| Minimal Skill Shape | Lists the required files (e.g., SKILL.md, bash scripts, or JSON schemas). |
Skill Learner Advanced Features
- Incremental Skill Evolution: Seamlessly patches existing Openclaw Skills instead of creating redundant documentation.
- Heuristic Filtering: Prevents "skill bloat" by requiring proof of a real, reusable pattern before recommending creation.
- Multi-level Implementation: Supports everything from lightweight local operating rules to complex, script-integrated automated workflows.
SKILL.md
Loading
Related Openclaw Skills
Get Taobao Month Sale
A specialized tool for extracting 30-day sales volume data from Taobao and Tmall listings via product IDs or URLs.
AI Upwork Proposal Writing & Job Tracking System
A privacy-first local AI assistant designed to filter Upwork job listings, draft high-conversion proposals, and maintain a historical ledger of applications.
Skill Discovery and Installation
A specialized skill designed to browse, find, and integrate new capabilities into your agent from the global repository.
Reading Assistant
An AI-powered personal reading assistant for managing EPUB libraries, generating chapter summaries, and tracking progress.
Nearby Car Parks
An automated geolocation skill designed to find and list parking facilities within a specified radius of a user's current coordinates.
Finnhub MCP Market Intelligence
A comprehensive Model Context Protocol server providing real-time financial data, sentiment analysis, and regulatory filings via the Finnhub API.
Featured*
