The Skill Creator provides a standardized framework and CLI tools for building, validating, and packaging modular extensions that enhance AI agent capabilities.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install skill-creator-2
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-creator-2 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 Creator?
This skill acts as a comprehensive blueprint for developers looking to extend agent functionality through Openclaw Skills. It transforms general-purpose models into domain experts by providing specialized procedural knowledge, tool integrations, and bundled resources. By following a structured three-level loading system—metadata, instruction body, and external resources—it ensures maximum efficiency within the AI context window.
The core philosophy of this framework is to challenge every piece of information added to the agent context. By using concise examples and maintaining appropriate degrees of freedom, developers can create robust Openclaw Skills that perform reliably across different LLM providers while minimizing token costs and maximizing deterministic behavior.
Skill Creator Use Cases
- Designing specialized workflows for multi-step domain tasks.
- Integrating custom APIs or proprietary file formats for agent use.
- Bundling reusable scripts and assets for repetitive agent actions.
- Optimizing token usage through progressive disclosure design patterns.
- Standardizing the development and distribution of new Openclaw Skills across a team.
How Skill Creator Works
- Analyze concrete usage examples to determine the required functionality and agent triggers.
- Plan reusable contents including Python/Bash scripts, reference documentation, and static assets.
- Initialize the skill directory structure using the provided CLI initialization script to create the required SKILL.md template.
- Draft the SKILL.md file with imperative instructions and precise YAML frontmatter metadata.
- Validate and package the folder into a distributable .skill file using the packaging utility.
- Test the skill in real-world scenarios and iterate based on performance feedback.
Skill Creator Setup
To initialize a new project within the Openclaw Skills ecosystem:
# Initialize a new skill directory
python scripts/init_skill.py my-new-skill --path ./skills
To package your skill for distribution:
# Validate and package the skill into a .skill file
python scripts/package_skill.py ./skills/my-new-skill
Skill Creator Data Schema & Taxonomy
Openclaw Skills follow a strict directory taxonomy to ensure agent compatibility and resource discovery:
| Component | Location | Description |
|---|---|---|
| SKILL.md | Root | Required. Contains YAML frontmatter (name, description) and instructions. |
| scripts/ | /scripts/ | Optional. Executable code (Python/Bash) for deterministic or repetitive tasks. |
| references/ | /references/ | Optional. Deep-dive docs, schemas, or policies loaded only when needed. |
| assets/ | /assets/ | Optional. Static files, boilerplate code, or templates for output generation. |
Skill Creator Advanced Features
- Progressive disclosure patterns to keep the context window lean by loading resources only on demand.
- Automatic validation of YAML frontmatter and directory structures during the packaging process.
- Domain-specific organization for handling multiple sub-tasks or providers within a single skill.
- Deterministic script execution to handle fragile, error-prone, or high-precision operations.
- Conditional detail loading to manage complex Openclaw Skills with multiple framework variations.
SKILL.md
Loading
Related Openclaw Skills
ClawConnect
A unified API bridge connecting AI agents to Twitter, Gmail, Slack, and Discord for seamless cross-platform communication.
RAGFlow Runbook
A comprehensive operational runbook designed for AI agents to deploy, monitor, and troubleshoot RAGFlow instances seamlessly.
Xiaomi Home Control
A high-performance local network controller for Xiaomi Home devices that bypasses cloud latency using miiocli.
Xiaomi Home Control
A specialized skill enabling direct local network control of Xiaomi Home devices using natural language and miiocli.
OpenClaw Configuration Manager
A comprehensive technical reference and management tool for the openclaw.json configuration file that powers the AI agent gateway.
Skill Vetter
A security-first vetting protocol designed to identify red flags and malicious code in AI agent skills before they are installed.
Featured*
