A meta-skill that empowers AI agents to architect, validate, and publish standardized skill modules on the ClawHub registry.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install eigenskill-builder
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 eigenskill-builder 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 Eigenskill Builder?
Eigenskill Builder is the definitive meta-skill designed to guide AI agents and developers in the creation of robust, standardized skills for the ecosystem. It establishes the foundational patterns for skill anatomy, YAML frontmatter specifications, and the directory conventions required for discovery on ClawHub. By implementing this skill, you ensure that every agent-led development task follows best practices for trigger conditions and negative boundaries, making Openclaw Skills more reliable and easier for agents to load autonomously.
This meta-skill acts as the blueprint for the entire skill-building lifecycle, covering everything from the initial directory structure to complex skill graph management. It ensures that content is actionable and machine-readable, providing a clear methodology for agents to teach other agents how to expand their capabilities within the framework of Openclaw Skills.
Eigenskill Builder Use Cases
- Creating new Openclaw Skills from scratch with standardized directory structures.
- Validating existing skills against strict quality standards before publishing to a registry.
- Designing precise skill descriptions that include essential trigger and exclusion boundaries for agent discovery.
- Setting up organized references directories for supplementary material without creating circular dependencies.
- Planning complex skill graphs for large libraries to manage dependencies between multiple Openclaw Skills.
How Eigenskill Builder Works
- Initialize the skill directory structure using the kebab-case naming convention and a mandatory SKILL.md file.
- Define the machine-readable identity using valid YAML frontmatter, ensuring specific trigger-rich descriptions for better searchability in Openclaw Skills.
- Establish negative boundaries within the documentation to clearly communicate what the skill does not cover, preventing agent logic conflicts.
- Organize supplementary material into a single-level references directory for focused, non-circular information retrieval.
- Implement automation scripts for validation to check for frontmatter accuracy, negative boundary presence, and directory depth.
- Utilize the ClawHub CLI to publish the validated skill, making it available for discovery and installation within the network of Openclaw Skills.
Eigenskill Builder Setup
To start building Openclaw Skills with the builder, set up your local environment and use the ClawHub CLI for lifecycle management.
# Search for existing templates
clawhub search "builder"
# Create the standard directory structure
mkdir -p my-new-skill/references
cd my-new-skill
# Run the validation script to ensure compliance
./scripts/validate.sh
# Publish the skill to the hub
clawhub publish ./ --dry-run
Eigenskill Builder Data Schema & Taxonomy
The Eigenskill Builder organizes Openclaw Skills using a strict file-based taxonomy to ensure agents can parse and execute instructions efficiently.
| Component | Path | Description |
|---|---|---|
| Skill Definition | SKILL.md |
The core file containing YAML frontmatter and operational instructions. |
| References | references/ |
A flat directory for detailed documentation that shouldn't clutter the main file. |
| Automation | scripts/ |
Contains validation and installation shell scripts for the skill. |
| Frontmatter | ---...--- |
Metadata block defining name, description, license, and dependencies. |
Eigenskill Builder Advanced Features
- Automated skill validation via custom shell scripts to enforce 10-point quality checklists before publishing Openclaw Skills.
- Support for Skill Graphs using YAML dependencies and wikilink references to create a navigable map of related agent capabilities.
- Semantic versioning (Semver) integration to manage breaking changes and feature updates across the skill library.
- Integrated description interview process to generate high-performance triggers and exclusions for superior agent discovery and loading logic.
SKILL.md
Loading
Related Openclaw Skills
Pixwith AI Media Generation
A comprehensive AI media generation skill for Openclaw Skills that enables text-to-video, image-to-video, and multi-model image creation via the Pixwith MCP server.
Vuact
Vuact is a powerful compatibility layer that enables seamless two-way interoperability between React and Vue 3 components.
Agent Orchestration
A technical framework for coordinating multiple AI models through quality control loops, task routing, and cost-optimized fallback chains.
Long Project Manager
A specialized skill for managing long-term project state through a structured, file-based workspace that survives chat resets.
SenseAudio Bedtime Radio
An AI-powered audio production skill that transforms simple keywords into fully produced, multi-character bedtime radio programs for children.
Xiaohongshu Publisher
A comprehensive automation skill for drafting, designing, and publishing high-engagement posts to Xiaohongshu using browser automation.
Featured*
