A comprehensive developer toolkit designed to create, evaluate, benchmark, and publish high-quality Openclaw Skills for AI agents.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install skill-factory
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-factory 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 Factory?
The Skill Factory is the definitive environment for developers to architect robust agent capabilities. It facilitates the entire development pipeline for Openclaw Skills, from initial scaffolding and pattern analysis to rigorous evaluation and community publishing via the clawhub registry. By utilizing standardized templates and automated testing scripts, it ensures that every skill meets high-performance benchmarks while maintaining a clean, manageable context footprint.
This utility operates across six specialized modes—Create, Eval, Improve, Benchmark, Analyze, and Synthesize—allowing developers to not only build from scratch but also to iterate based on data-driven feedback. It serves as a centralized hub for managing the lifecycle of Openclaw Skills, ensuring they are trigger-ready and technically sound before deployment.
Skill Factory Use Cases
- Generating new Openclaw Skills from scratch using standardized scaffolds and best practices.
- Running automated evaluations to measure the quality and trigger accuracy of existing Openclaw Skills.
- Benchmarking different versions of a skill to perform blind A/B testing and performance delta analysis.
- Extracting reusable architectural patterns from established Openclaw Skills to synthesize new functionality.
- Packaging and publishing verified capabilities to the community registry for wider ecosystem use.
How Skill Factory Works
- Analyze existing patterns using the built-in scripts to identify common triggers and tool requirements.
- Initialize the skill directory structure, including the core SKILL.md file and resource folders for scripts and references.
- Author the technical specification using progressive disclosure rules to optimize agent context management.
- Execute automated eval suites to calculate pass rates based on specific functional assertions and prompts.
- Iterate on the logic and metadata to reach a minimum quality bar of 0.80 for publishing.
- Synchronize and publish the finalized assets to the community registry using the integrated CLI tools.
Skill Factory Setup
To begin building your own Openclaw Skills, use the bundled initialization script within your workspace:
# Initialize a new skill directory
python3 scripts/init_skill.py <skill-name> --path ~/.openclaw/workspace/skills/ --resources scripts,references
# Validate and package the skill
python3 scripts/package_skill.py ~/.openclaw/workspace/skills/<skill-name>
# Publish to the community registry
clawhub login
clawhub publish ~/.openclaw/workspace/skills/<skill-name> --version 1.0.0
Skill Factory Data Schema & Taxonomy
Skill Factory organizes its assets into a strict hierarchy to ensure compatibility across Openclaw Skills environments:
| File/Folder | Purpose |
|---|---|
SKILL.md |
The primary instruction set containing frontmatter metadata and usage recipes. |
scripts/ |
Helper Python or Bash scripts bundled for complex logic execution. |
evals/ |
Directory containing evals.json test cases and historical run results. |
_meta.json |
Auto-populated publishing metadata including versioning and slug identifiers. |
references/ |
Supporting documentation, API schemas, and extracted pattern libraries. |
Skill Factory Advanced Features
- Pattern Analysis: Scans multiple directories to extract reusable tool patterns and trigger phrases from existing Openclaw Skills.
- Blind A/B Benchmarking: Compares two versions of a skill without labels to determine the objective winner based on output quality.
- Progressive Disclosure: Efficiently manages the agent context window by separating core triggers from on-demand verbose resources.
- Automated Evaluation Framework: Tracks version history and pass rates across multiple iterations to ensure zero regressions.
- Registry Integration: Direct support for clawhub commands to sync and version-control your entire library of Openclaw Skills.
SKILL.md
Loading
Related Openclaw Skills
MLB Daily Scores
A dedicated MLB companion that fetches daily game recaps, upcoming previews, and injury reports for your favorite baseball team.
neolata-mem
A high-performance, graph-native memory engine for AI agents featuring hybrid search, biological decay, and Zettelkasten-style linking.
YouTube AI Editor
A comprehensive automation tool for downloading videos, generating subtitles, analyzing SEO content, and creating character-consistent AI thumbnails.
WebNovel Serial Pipeline
A professional publishing pipeline for Quartz-hosted Korean web-novels that automates image optimization, linting, and deployment.
xdotool-control
A powerful automation skill for controlling Linux desktop interfaces, mouse movements, and keyboard inputs without a browser relay.
Limitless Life Logs
An AI-powered skill to search, summarize, and extract actionable tasks from your Limitless AI pendant recordings.
Featured*
