A framework and toolkit for developing, initializing, and packaging modular extensions for AI agents.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install create-agent-skills
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 create-agent-skills 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 Designer Agent?
The Skill Designer Agent provides a structured methodology for creating Openclaw Skills, which serve as specialized onboarding guides for AI agents. These skills transform general-purpose models into domain experts by providing them with procedural knowledge, API integrations, and specialized schemas. By focusing on modularity and efficiency, the framework ensures that agents remain performant within limited context windows.
At its core, this skill teaches developers how to balance freedom and specificity when designing agent workflows. Whether you are building low-freedom deterministic scripts or high-freedom text-based instructions, the Skill Designer Agent ensures that your Openclaw Skills are organized, discoverable, and ready for distribution.
Skill Designer Agent Use Cases
- Standardizing agent behavior for specific business domains or technical workflows.
- Creating reusable tool integrations to handle complex file formats like PDFs or DOCX.
- Developing modular knowledge bases that agents can reference on-demand.
- Building custom Openclaw Skills to automate repetitive coding or data processing tasks.
- Managing complex multi-step procedures with progressive disclosure of information.
How Skill Designer Agent Works
- Identify the core purpose and concrete usage examples of the new skill.
- Map out the required resources including scripts for automation and reference docs for domain expertise.
- Execute the initialization script to generate a standardized directory structure.
- Define the skill's triggering mechanism in the SKILL.md YAML frontmatter.
- Populate the body with imperative instructions and link to external reference files.
- Run the packaging utility to validate the structure and generate a distributable file.
Skill Designer Agent Setup
To initialize a new project within the ecosystem of Openclaw Skills, run the following command:
scripts/init_skill.py <skill-name> --path <output-directory>
After adding your content, use the packaging tool to validate and create the final distribution file:
scripts/package_skill.py <path/to/skill-folder>
Skill Designer Agent Data Schema & Taxonomy
The architecture for Openclaw Skills is designed for efficiency and clarity:
| Component | Description | Requirement |
|---|---|---|
| SKILL.md | Contains YAML metadata and core instructions | Required |
| scripts/ | Executable Python or Bash scripts for deterministic tasks | Optional |
| references/ | Deep-dive documentation or schemas loaded on-demand | Optional |
| assets/ | Static resources like templates, icons, or boilerplate code | Optional |
Skill Designer Agent Advanced Features
- Three-level progressive disclosure system to optimize context window usage.
- Automated validation during the packaging phase to ensure metadata and directory integrity.
- Support for domain-specific reference patterns to avoid loading irrelevant context.
- Deterministic script execution for fragile or repetitive code operations.
SKILL.md
Loading
Related Openclaw Skills
ClawChain RPC Client
An RPC client connecting EvoClaw agents to the ClawChain L1 blockchain for on-chain identity, reputation, and tokenomics.
ClawChain Contributor
An essential skill for agents to contribute code, documentation, and architecture decisions to the ClawChain blockchain ecosystem.
clangd LSP
A professional-grade C/C++ language server providing deep code intelligence, real-time diagnostics, and automated formatting.
browser-use
An advanced AI-driven browser automation tool that leverages LLMs to navigate, interact with, and extract structured data from any website using a real Chromium instance.
FearHarvester Skill
An autonomous sentiment-based trading agent that executes DCA strategies into BTC and ETH during periods of extreme market fear.
gopls-lsp
A powerful Go language server integration providing real-time code intelligence, diagnostics, and refactoring support for Go developers.
Featured*
