A powerful utility for stripping web page clutter to extract clean, readable article content and structured metadata.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install defuddle
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 defuddle 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 Defuddle Web Cleaner?
Defuddle Web Cleaner is a specialized tool designed to solve the problem of 'web noise.' By leveraging the Defuddle parser, this skill identifies and removes non-essential elements like navigation bars, sidebars, advertisements, and social buttons. It is an essential addition to the library of Openclaw Skills for developers and researchers who need high-quality, distraction-free text.
Whether you are processing raw HTML or a live URL, this skill transforms messy web data into clean article content. It is particularly effective for preparing data for LLM consumption, ensuring that models focus on the actual article text rather than site boilerplate.
Defuddle Web Cleaner Use Cases
- Clipping web articles directly into Obsidian or other personal knowledge management systems.
- Building high-quality research datasets by extracting text from multiple news sources.
- Cleaning web pages to reduce token usage and improve accuracy in LLM summarization tasks.
- Converting complex web layouts into simple, portable Markdown files.
How Defuddle Web Cleaner Works
- Input Detection: The skill identifies if the provided input is a URL, raw HTML string, or unstructured web text.
- HTML Retrieval: For URLs, the tool fetches the source HTML; for raw inputs, it prepares the content for parsing.
- Defuddle Parsing: The core engine runs the Defuddle algorithm to isolate the 'readable' container of the page.
- Metadata Extraction: Specific fields like author, title, and publish date are parsed from the headers and OpenGraph tags.
- Markdown Conversion: The cleaned HTML is converted into valid Markdown syntax.
- Response Delivery: The system returns either a formatted text block or a structured JSON object.
Defuddle Web Cleaner Setup
To integrate this capability into your environment, ensure you have the Openclaw Skills runner configured. Install the package using your preferred package manager:
npm install @openclaw/skill-defuddle-web-cleaner
Configure your API keys if the environment requires remote fetching, and invoke the skill by passing a url or html string to the handler.
Defuddle Web Cleaner Data Schema & Taxonomy
The skill returns data in two primary formats depending on your request. The structured JSON schema is detailed below:
| Property | Type | Description |
|---|---|---|
title |
String | The main headline of the article |
author |
String | The credited creator of the content |
site |
String | The name of the source website |
description |
String | A brief summary or meta-description |
published |
ISO Date | The date the article was posted |
content |
HTML | The cleaned HTML fragment |
contentMarkdown |
String | The final converted Markdown text |
Defuddle Web Cleaner Advanced Features
- Multi-input support: Seamlessly switches between processing live URLs and static HTML files.
- LLM Optimization: Specifically tuned to produce clean Markdown that maximizes context window efficiency for AI agents.
- Metadata Taxonomy: Automatically maps various web standards (Schema.org, Dublin Core, OpenGraph) into a unified JSON structure.
- Integration Friendly: Built to work natively within broader research workflows using Openclaw Skills.
SKILL.md
Loading
Related Openclaw Skills
ITSM Ticket Analyzer
An intelligent analysis tool for BlueKing ITSM tickets that provides processing suggestions, trend reports, and SLA risk monitoring.
Numerology Calculator
A professional numerology engine providing deep name and birth date analysis using Pythagorean and Chaldean systems.
Tianqi Weather Skill
A specialized weather query skill optimized for Chinese language users providing real-time conditions and structured forecasts.
Amazon Listing Optimizer
A zero-cost, high-performance toolset for analyzing Amazon product listings and performing deep keyword research without paid APIs.
Source Research
A standardized framework for discovering, evaluating, and automating high-quality information source pools to build reusable research artifacts.
Brave API Free Search (Secure)
A 100% free and secure localhost-only web search alternative to the Brave API for OpenClaw agents.
Featured*
