OpenClaw Skills Directory

A Curated Directory of OpenClaw Skills

Discover 1,700+ verified skills to power up your AI agent. We filter out the noise so you can find safe, high-quality OpenClaw skills to automate your workflows.

1,700+OpenClaw skills indexed
Updated daily

How to Install OpenClaw Skills

Choose the installation method that works best for you

Recommended

ClawHub CLI

Fast and simple installation of OpenClaw skills using the official ClawHub CLI tool

Note: As you probably know, they keep renaming things. This reflects the current official docs.

Install any skill folder in one shot:
npx clawhub@latest install <skill-slug>

Manual Installation

Advanced

Copy the OpenClaw skills folder to one of these locations

TargetPath
Global~/.openclaw/skills/
Workspace<project>/skills/

Priority: Workspace > Local > Bundled

Alternative Method

Easiest

Paste the skill's GitHub repository link directly into your assistant's chat and ask it to use it. The assistant will handle the OpenClaw skills setup automatically in the background.

How OpenClaw Skills Power Your Agent

Understanding the architecture behind autonomous OpenClaw skills

It starts with a SKILL.md

Each of the OpenClaw skills is defined by a simple Markdown file. It tells the agent what it can do (e.g., 'Manage Calendar') and how to do it (e.g., run `python calendar.py`).

The LLM Decides

When you ask OpenClaw to 'Schedule a meeting,' the LLM analyzes your request, scans available OpenClaw skills, and intelligently picks the right tool for the job.

Local Execution

The agent executes OpenClaw skills scripts specifically on your local machine—keeping your data private and your actions direct. No middleman servers involved.

Why Choose OpenClaw Skills?

OpenClaw’s public registry hosts 3,000+ community-built skills as of February 2, 2026. Every one of the OpenClaw skills listed here is pulled directly from that registry; this is simply a curated, categorized subset to make discovery easier.

From the full registry, we selected 1,705+ OpenClaw skills by excluding crypto/blockchain/DeFi content, spam or bulk-generated entries, duplicates, and items with non-English descriptions. We also attempted to remove skills that appear risky or harmful (e.g., abuse, bypass, or fraud-related), though we can’t guarantee we caught everything.

Being included in this list does not mean a skill is safe or trustworthy. We recommend reviewing a skill before installing it, and using tools like Claude Code or Codex to inspect it for potentially harmful behavior.

If you believe a skill was excluded in error or categorized incorrectly, please contact us at [email protected]. We may have made mistakes.

Frequently Asked Questions

Common questions about OpenClaw Skills and the ecosystem

General

Security & Usage