A standardized framework for building, validating, and sharing portable instructions and capabilities across diverse AI agent platforms.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agentskills-io
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 agentskills-io 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 Agent Skills (agentskills-io)?
Agent Skills (agentskills-io) provides a universal specification for defining AI agent capabilities through a structured markdown and metadata approach. By following this open standard, developers can create skills that work seamlessly across Claude Code, Cursor, GitHub Copilot, and other LLM-powered environments. This framework allows for high-quality discovery by agents and ensures that instructions remain modular and maintainable.
Integrating your custom tools with Openclaw Skills using this specification allows for a platform-agnostic developer experience. The standard emphasizes strict token budgets and imperative language, ensuring that agents can quickly parse and execute the necessary logic without exhausting context windows or introducing ambiguity.
Agent Skills (agentskills-io) Use Cases
- Creating new portable skills for AI agents with standardized discovery metadata.
- Validating skill structure and frontmatter to ensure compatibility with major IDEs.
- Converting existing technical documentation into actionable agent instructions.
- Sharing agent capabilities across different environments using symlink integration.
- Managing large repositories of agentic workflows with automated batch validation.
How Agent Skills (agentskills-io) Works
- Initialize a skill directory containing a mandatory SKILL.md file with standardized YAML frontmatter.
- Define discovery metadata including a name and a specific Use when... description to help the agent activate the skill.
- Author the skill body using imperative instructions, concrete examples, and troubleshooting steps.
- Use the skills-ref CLI tool to validate the directory structure, metadata schema, and token counts.
- Deploy the skill by placing it in a recognized directory or symlinking it to environments like Openclaw Skills for instant access.
Agent Skills (agentskills-io) Setup
To get started with validating and managing your skills, install the official toolkit using uv:
# Install the validation tool
uv tool install git+https://github.com/agentskills/agentskills#subdirectory=skills-ref
# Create a new skill
mkdir my-skill && touch my-skill/SKILL.md
# Run a validation check
skills-ref validate ./my-skill
Agent Skills (agentskills-io) Data Schema & Taxonomy
The skill standard follows a strict directory and file structure to ensure portability:
| Path | Requirement | Description |
|---|---|---|
SKILL.md |
Required | Contains frontmatter and core instructions (limit 5000 tokens) |
scripts/ |
Optional | Contains executable code or automation logic |
references/ |
Optional | Detailed documentation that exceeds the SKILL.md token budget |
assets/ |
Optional | Static files, templates, or non-executable resources |
The YAML frontmatter must include a name (lowercase alphanumeric-hyphens) and a description (under 1024 characters).
Agent Skills (agentskills-io) Advanced Features
- Symlink sharing for cross-platform synchronization between Openclaw Skills, Claude Code, and Cursor.
- Automated semver version bumping for plugin-based skill deployments.
- Progressive disclosure patterns using references to keep the primary prompt context lean.
- Batch validation scripts for maintaining integrity across massive skill repositories.
- Subagent support via dedicated agents directories for complex multi-agent orchestration.
SKILL.md
Loading
Related Openclaw Skills
Adversarial Coach
A dialectical AI reviewer that validates implementation completeness against requirements through rigorous adversarial logic.
Virtuoso Support Agent
An advanced AI-driven assistant designed for technical support, database administration, and automated RDF Views generation for OpenLink Virtuoso Servers.
Wikidata Query Skill
The Wikidata Query Skill empowers AI agents to translate plain English questions into complex SPARQL queries to extract structured data from Wikidata.
DBpedia Query Skill
A powerful integration that transforms natural language questions into complex SPARQL queries to extract and visualize structured data from DBpedia.
AWS AgentCore + LangGraph
A production-ready framework for deploying stateful, multi-agent LangGraph systems on AWS Bedrock AgentCore infrastructure.
BAML Code Generation
A specialized skill for generating type-safe BAML code to handle structured LLM outputs, classification, and RAG workflows with 95% compilation success.
Featured*
