ClawhHub Skills Marketplace: Best OpenClaw Skills to Install in 2026

ByL.B. Kaelin

The ClawhHub skills marketplace makes it easy to find and install the best OpenClaw skills in 2026. The fastest way to extend what OpenClaw can do takes about sixty seconds: install a skill from clawhub.ai. Skills are self-contained instruction modules that give your agent new capabilities — weather forecasts, smart home control, GitHub pull request reviews, security audits, and more. This guide covers the top OpenClaw skills to install in 2026, how to find them on ClawhHub, and what to check before installing community-contributed skills.

What ClawhHub Is and How Skills Work

ClawhHub (available at clawhub.ai) is the community skills marketplace for OpenClaw. Users share, rate, and install skills — self-contained instruction modules that give OpenClaw agents specialized capabilities. Think of it as an app store for agent behaviors, with the important difference that skills are plain-text instruction files the agent reads and follows, not compiled binaries.

Each skill is a directory containing a SKILL.md file and optional supporting scripts or reference files. The SKILL.md file tells the agent what the skill does and when to activate it, based on a description trigger. When you ask OpenClaw something that matches a skill’s trigger, the agent reads the skill and uses its instructions to handle your request.

Skills on ClawhHub fall into several categories: productivity, communication, data analysis, media, security, finance, home automation, research, coding, and writing. Hundreds of skills are available across these categories, and the marketplace continues to grow as more users publish their work.

How to Install a Skill from ClawhHub

Installing a skill takes roughly one minute. Here is the step-by-step process:

  1. Visit clawhub.ai and find a skill. Browse by category or search for a specific capability. Each skill listing shows a description, rating, author, and download link.
  2. Download the skill directory. ClawhHub packages each skill as a directory containing the SKILL.md file and any supporting files.
  3. Place the directory in your OpenClaw skills/ folder. Copy the entire skill directory into the skills/ directory where OpenClaw is installed. The exact path depends on your setup — for a standard OpenClaw installation, this is typically ~/.openclaw/skills/ or alongside your openclaw.json configuration file.
  4. Restart OpenClaw or reload skills. The agent scans the skills directory at startup. Some setups support a skills reload command without a full restart.
  5. Test the skill. Ask OpenClaw something the skill should handle. If the skill’s description trigger matches your request, the agent will use it.

That is the entire process. No package manager, no dependencies to install, no API keys to configure for most skills. Some skills — like weather — work immediately with zero configuration. Others may require an API key or account setup, which the skill’s documentation will specify.

Official OpenClaw Skills: The Safe Starting Point

The skills that ship with OpenClaw are maintained by the core development team. They are safe to install, well-documented, and cover the most common use cases. Every OpenClaw user should start with these before exploring community contributions.

weather

Gets current weather conditions and forecasts via wttr.in and Open-Meteo. No API key needed. Simply ask “What is the weather in Tokyo?” or “Will it rain tomorrow in Berlin?” and the skill handles the rest. This is arguably the skill with the highest immediate usefulness-to-effort ratio in the entire marketplace.

taskflow

Durable task management for complex multi-step automation. Supports child-task linkage, waiting states, revision tracking, and error recovery. Essential if you build any non-trivial automation with OpenClaw — fetching data, transforming it, and publishing results as a multi-step pipeline.

healthcheck

Security hardening and posture review for the machine running OpenClaw. Checks SSH configuration, firewall rules, open ports, pending package updates, and known CVEs. Produces a risk report with specific remediation steps. Essential for anyone self-hosting OpenClaw on a VPS or publicly accessible machine.

node-connect

Diagnoses and fixes pairing failures between OpenClaw and its iOS/Android companion apps. If you are setting up the mobile app and see “pairing required,” “unauthorized,” or “bootstrap token invalid,” this skill walks through the most common causes and solutions. It resolves the majority of “can’t connect” issues.

skill-creator

A meta-skill: it helps you write better skills by following the SKILL.md specification. When you run into questions about skill format, triggering, or best practices, this skill gives you the answers. Use it when creating your first custom skill.

taskflow-inbox-triage

An example pattern built on top of taskflow that demonstrates how to handle incoming messages based on intent — some routed for immediate attention, others held for later response, others rolled into a summary. Useful as a starting template for inbox automation.

slack

Controls Slack from OpenClaw: reacting to messages, pinning and unpinning items, sending messages to channels or DMs. Useful for teams that want their OpenClaw instance participating in Slack workflows.

These official skills are maintained and tested by the OpenClaw team. They do not run arbitrary shell commands without your knowledge, and they do not access files outside the designated workspace. They are the safest place to start building your skill collection.

Best Community Skills to Install in 2026

The community-sourced skills on ClawhHub extend OpenClaw into areas the official set does not cover. Some require additional services or API keys; others work out of the box. Here are the ten most valuable community skills available in 2026.

news-digest

Pulls headlines from RSS feeds and summarizes them. You configure the feeds you want to monitor, and the skill fetches new items on a cron schedule. Good for building a personal news briefing that runs every morning, or for monitoring industry-specific sources without checking a dozen websites.

Setup: Define your RSS feed URLs in a configuration file. The skill handles fetching and summarizing. You can set a cron interval via OpenClaw’s scheduling system.

home-assistant-control

Connects OpenClaw to Home Assistant for smart home control via natural language. Turn on lights, adjust thermostats, check sensor status, or run automations by asking OpenClaw directly. Requires a running Home Assistant instance with a long-lived access token.

Setup: Provide your Home Assistant URL and a long-lived access token from your Home Assistant profile page. The skill uses the Home Assistant REST API and does not require WebSocket or websocket-specific configuration.

github-pr-review

Automated pull request review for GitHub repositories. When a pull request is opened or updated, this skill checks for common issues — missing tests, large diffs, hardcoded credentials, syntax errors — and posts review comments. Designed for solo developers and small teams who want automated review without paying for GitHub Copilot or similar services.

Setup: Requires a GitHub personal access token with repo scope. Configure the skill with the repositories you want it to monitor. Webhook setup is handled by the skill’s instructions.

calendar-sync

Reads events from Google Calendar or Apple Calendar and can add new ones. Useful for scheduling assistance: “What does my calendar look like tomorrow?” or “Add a meeting at 2 PM on Friday.” Works with the calendar provider’s API and requires OAuth setup for Google Calendar or local calendar access for Apple Calendar.

Setup: Google Calendar requires OAuth credentials (client ID and secret) from the Google Cloud Console. Apple Calendar access depends on macOS Calendar permissions and may require additional configuration on non-macOS systems.

rss-monitor

Monitors RSS and Atom feeds for new content and alerts you when specific keywords or topics appear. More targeted than news-digest — good for competitive intelligence, tracking specific companies or products, or monitoring job boards.

Setup: Requires a list of feed URLs and keyword patterns. The skill can be configured to run on a cron schedule and notify via Slack, email, or a desktop notification.

code-executor

Lets OpenClaw execute code snippets in a sandboxed environment. Supports Python, JavaScript, Go, and Rust. Useful for quick calculations, testing code ideas, or running scripts without switching to a terminal tab. Runs in a Docker container by default to limit security risk.

Setup: Docker must be installed and the OpenClaw user must have permission to run Docker containers. The skill pulls a sandbox image on first use.

web-scraper

Fetches and extracts structured data from web pages. Unlike basic web fetching, this skill can handle pagination, form submissions, and JavaScript-rendered content (via Playwright). Designed for data collection tasks like scraping product listings, job postings, or documentation.

Setup: Requires Playwright to be installed for JavaScript-rendered pages. Basic scraping works without additional tools.

file-organizer

Scans directories and organizes files by type, date, or custom rules. Moves files into categorized folders, renames them based on patterns, and can archive old files. Useful for keeping download folders or project directories clean without manual effort.

Setup: Define the directories and rules in a configuration file. The skill runs on demand or on a schedule.

financial-tracker

Parses bank statement exports (CSV format) and categorizes spending. Generates monthly spending reports by category, tracks trends over time, and can alert you when spending in a category exceeds a threshold. All processing happens locally — no financial data leaves your machine.

Setup: Export your bank statements as CSV files and place them in a directory the skill can access. Define category mapping rules in a configuration file.

docker-manager

Manages Docker containers, images, volumes, and networks through OpenClaw. Start, stop, inspect, and monitor containers with natural language. Also checks for available image updates and reports container health status.

Setup: Requires the Docker socket or API endpoint to be accessible from the OpenClaw host. The skill uses the Docker SDK or shell commands depending on configuration.

Skill Vetting: What to Check Before Installing from the Community

ClawhHub has a community rating system, but it does not have a formal security review process. Anyone can publish a skill. Before installing any community skill, run through this checklist:

  1. Read the SKILL.md file fully. Open it and read every line. A well-written SKILL.md explains exactly what the skill does, what commands it runs, and what data it accesses. If the documentation is sparse, unclear, or contains grammatical errors that obscure meaning, treat that as a warning sign.
  2. Check if the skill runs shell commands. Search the SKILL.md and any associated scripts for shell execution patterns. Skills that run arbitrary shell commands can do anything the OpenClaw user can do — including reading, writing, and deleting files. Only install skills from authors you trust.
  3. Verify it does not access files outside the workspace. A skill should operate within the OpenClaw workspace directory. If a skill references paths like /etc/, /home/, /root/, or uses relative paths like ../../, it may be accessing files it should not. Check the skill’s file access patterns carefully.
  4. Check the author’s reputation. Does the author have multiple skills on ClawhHub? Positive ratings? A history of responding to issues? An anonymous or newly-created account with no history should be treated with higher scrutiny.
  5. Review external service connections. If the skill connects to an external API, verify that it uses HTTPS and that credentials are configured through environment variables or a secure config file, not hardcoded in the skill.
  6. Look at the skill’s dependencies. Some skills require npm packages, Python libraries, or system tools. Ensure these are documented and that you are comfortable with the security posture of those dependencies.
  7. Test in a sandbox first. If possible, test a new community skill in an isolated environment or a non-production OpenClaw instance before running it with sensitive data or access.
  8. Read the OpenClaw security documentation. The official OpenClaw documentation covers security best practices for skills, including sandboxing options and permission models available in your version.

The official skills that ship with OpenClaw (weather, taskflow, healthcheck, node-connect, skill-creator, taskflow-inbox-triage, slack) are maintained by the OpenClaw team. They are safe to install and do not require this level of scrutiny. Apply the checklist above only to community-contributed skills from ClawhHub.

Skills That Work Well Together: Combinations and Synergies

Some skills become significantly more powerful when combined. Here are the most useful pairings and combinations.

healthcheck + node-connect

When setting up OpenClaw on a new machine, run healthcheck first to ensure the system is properly secured, then run node-connect to pair mobile devices. This sequence prevents common security issues — like having SSH exposed with password authentication — before you connect your mobile app.

taskflow + github-pr-review

Use taskflow to create a durable pipeline that monitors pull requests, runs github-pr-review on each new PR, collects the results, and posts them to a Slack channel. If the review fails at any step, taskflow can retry or alert you. This combination gives solo developers a CI-style review pipeline without a full CI system.

news-digest + calendar-sync

Start your day with a news digest from your RSS feeds, then have calendar-sync check your day’s schedule and summarize it. Two sequential requests that together give you a complete morning briefing. Can be combined into a single scheduled task using OpenClaw’s cron capabilities.

docker-manager + healthcheck

Healthcheck reports on host security, but docker-manager can check the security posture of your running containers — exposed ports, outdated images, unhealthy containers. Together they provide a more complete security picture than either alone.

skill-creator + web-scraper

Use web-scraper to gather data or examples from the web, then feed that into skill-creator to help you build a new skill based on real patterns. This combination accelerates the process of creating your own skills by providing concrete reference material.

financial-tracker + taskflow

Schedule financial-tracker to run weekly via taskflow’s durable execution: parse the latest CSV exports from your bank, categorize transactions, generate a spending report, and post the summary to a private Slack channel. Taskflow handles retries if the CSV file is not yet available or if the bank’s export format changes.

How to Write Your Own Skill and Submit to ClawhHub

Creating a custom skill for OpenClaw is straightforward. Every skill is a directory containing a SKILL.md file that follows the OpenClaw skill specification. The OpenClaw team also publishes the skill-creator skill, which serves as a guided assistant for writing skills that conform to the specification.

The basic process for creating a skill:

  1. Create a directory with a descriptive name for your skill.
  2. Write a SKILL.md file that includes a description telling OpenClaw when to activate the skill, along with the instructions the agent should follow.
  3. Add any supporting files your skill needs — reference documents, shell scripts, or configuration templates — in the same directory.
  4. Test your skill by placing it in your OpenClaw skills directory and interacting with the agent.
  5. Submit to ClawhHub by creating an account at clawhub.ai and following the submission process.

The skill-creator skill is the best resource for your first custom skill. Install it from the official set, then ask OpenClaw for help designing your skill format. It follows the SKILL.md specification and will catch common mistakes before you submit.

For a detailed walkthrough of the SKILL.md format and structure, see the related article on skill creation at the link in the Sources section below.

Sources

  • ClawhHub Marketplace: clawhub.ai
  • OpenClaw official documentation: referenced in-memory for skill installation paths, skill directory structure, and reload capabilities.
  • Skill format specification: The SKILL.md format is documented in the OpenClaw project README and developer guides.

Related Reading

Similar Posts