A professional framework for developing, validating, and packaging modular capabilities to extend AI agent functionality through specialized knowledge and tools.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install creator
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 creator 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?
The Skill Creator is a comprehensive development framework designed to transform a general-purpose AI into a specialized agent equipped with procedural knowledge. By following the standards for Openclaw Skills, developers can create modular packages that bundle domain expertise, multi-step workflows, and deterministic scripts. This ensures that agents can handle complex, repetitive tasks with a level of precision that general models cannot achieve alone.
At its core, this skill emphasizes context efficiency through a progressive disclosure design. It allows developers to manage the AI's context window by separating high-level triggers from deep reference material and executable code. By utilizing the Skill Creator, you ensure that your Openclaw Skills are professional, validated, and optimized for high-performance agentic workflows.
Skill Creator Use Cases
- Developing new specialized capabilities to extend AI agent functionality.
- Standardizing company-specific logic and schemas into reusable Openclaw Skills.
- Creating deterministic scripts for fragile operations like PDF manipulation or API interactions.
- Organizing large documentation sets into searchable references for AI consumption.
- Packaging boilerplate code and brand assets for automated content generation.
How Skill Creator Works
- Analyze concrete usage examples to understand the specific functionality the agent needs to support.
- Plan the required reusable resources, including executable scripts, reference markdown files, and static assets.
- Initialize the directory structure using automated CLI tools to ensure compliance with the Openclaw Skills architecture.
- Draft the SKILL.md file, focusing on precise YAML frontmatter for accurate triggering and imperative instructions for execution.
- Run validation scripts to check for metadata accuracy, file organization, and naming conventions.
- Package the finalized directory into a distributable .skill file for deployment and sharing.
Skill Creator Setup
To start building Openclaw Skills, you must first initialize your environment and use the provided automation scripts. Use the following commands to manage the lifecycle of your skill:
Initialize a new skill directory:
scripts/init_skill.py <skill-name> --path <output-directory>
Validate your skill structure and metadata before distribution:
scripts/quick_validate.py <path/to/skill-folder>
Package the skill into a distributable file:
scripts/package_skill.py <path/to/skill-folder>
Skill Creator Data Schema & Taxonomy
The architecture for Openclaw Skills is designed for modularity and context optimization. The following table describes how data is organized within a skill package:
| Component | Location | Description |
|---|---|---|
| Metadata | SKILL.md (Frontmatter) | Contains the name and description; used as the primary trigger for the agent. |
| Instructions | SKILL.md (Body) | Markdown-formatted procedural guidance loaded only when the skill is active. |
| Scripts | /scripts | Executable Python or Bash code for deterministic, token-efficient operations. |
| References | /references | Documentation, schemas, and policies loaded into context on an as-needed basis. |
| Assets | /assets | Static templates, logos, and boilerplate files used in the agent's output. |
Skill Creator Advanced Features
- Progressive Disclosure: A three-level loading system that keeps the context window lean by only loading metadata, body content, or resources when required.
- Deterministic Execution: Support for external scripts that ensure reliable output for fragile or repetitive tasks without relying on LLM code generation.
- Domain-Specific References: Capability to organize complex Openclaw Skills into sub-domain files to prevent irrelevant context bloat.
- Automated Packaging: Integrated validation and compression scripts that prepare skills for immediate distribution.
- Conditional Degrees of Freedom: Guidelines for balancing agent autonomy with strict procedural guardrails based on task complexity.
SKILL.md
Loading
Related Openclaw Skills
Signup Lead Integration
A specialized skill for programmatically capturing and transmitting lead details to the AgenticCreed system.
IdentityGram Signin
A specialized authentication skill designed to facilitate secure user sign-in to the IdentityGram platform via its gateway API.
Signup Lead for AgenticCreed
Automates the creation of signup leads in the AgenticCreed system via a secure public HTTP endpoint.
Markdown Validator
A high-performance CLI tool designed to identify broken internal links and maintain documentation integrity within Markdown files.
Volcengine Image Generation
A specialized skill for generating high-quality images from text descriptions using Volcengine's Ark API.
Volcengine Video Generate
A powerful CLI utility for generating AI videos from text descriptions and initial frame references using Volcengine's API.
Featured*
