A robust extraction tool that transforms any webpage URL into clean Markdown, plain text, or structured HTML for easy processing.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install felo-web-extract
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install felo-web-extract using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
The Felo Web Extract skill is a high-performance utility designed to bridge the gap between complex web structures and clean data formats within Openclaw Skills. By leveraging the Felo Web Extract API, it allows developers to bypass the traditional hurdles of web scraping, such as handling dynamic content or stripping away boilerplate code. This skill is essential for anyone building AI-driven workflows that require high-quality, readable content from the web.
Whether you need to convert a technical blog post into Markdown for documentation or extract specific data points using CSS selectors, this tool provides a reliable and scalable solution. It integrates seamlessly into the Openclaw Skills ecosystem, offering both a simplified CLI interface and direct API access for advanced automation tasks.
To get started, you must first obtain an API key from felo.ai via the Settings -> API Keys menu. Once you have your key, configure your environment variable:
# Linux/macOS
export FELO_API_KEY="your-api-key-here"
# Windows PowerShell
$env:FELO_API_KEY="your-api-key-here"
You can then run the skill via the bundled script:
node felo-web-extract/scripts/run_web_extract.mjs --url "https://example.com"
The skill organizes extracted data based on the requested format. The underlying API response follows this structure:
| Property | Description |
|---|---|
code |
Integer status code (0 indicates success). |
message |
String confirmation message. |
data.content |
The extracted string containing the Markdown, HTML, or Text. |
When using the Openclaw Skills CLI, users can pass the --json flag to receive the full metadata object, otherwise, only the raw content string is returned for easier piping into other tools.
target-selector to pinpoint exactly which part of the DOM to extract.wait-for-selector option ensures that JavaScript-heavy pages are fully loaded before extraction.fast for speed or fine for more thorough page parsing.Loading
A real-time web search skill that bridges the gap between static LLM knowledge and live internet data.

An automated skill that uses xAI Grok Imagine to edit a character reference image and distribute the result across messaging platforms via OpenClaw.

An automated assistant for troubleshooting and implementing Feishu (Lark) API integrations, authentication flows, and bot communications.

A browser automation skill for the Xiaohongshu platform designed to automate the publishing of long-form notes and content.

A high-performance web extraction tool that converts webpage content into clean, AI-ready formats like Markdown and plain text.

A high-performance tool for fetching and extracting YouTube video subtitles and transcripts in various languages with optional timestamp support.
