Extract clean, clutter-free markdown from web pages to optimize token usage and research accuracy.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install defuddle-obsidian
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-obsidian 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-to-Markdown Parser?
Defuddle is a specialized skill designed for developers and researchers who need to ingest web content without the noise of ads, navigation bars, or scripts. By leveraging the Defuddle CLI, this skill transforms complex URLs into readable Markdown, making it an essential component of the Openclaw Skills library for those using AI agents for data analysis. It provides a significant advantage over standard fetch tools by reducing token consumption and focusing purely on relevant text content.
Defuddle Web-to-Markdown Parser Use Cases
- Summarizing long-form articles or blog posts without distraction.
- Importing clean documentation into knowledge bases like Obsidian.
- Reducing input tokens for LLM-based analysis of web pages.
- Automating the collection of metadata from multiple URLs.
How Defuddle Web-to-Markdown Parser Works
- The user provides a target URL to the AI agent for analysis.
- The skill invokes the Defuddle CLI parse command using the --md flag to prioritize readable text.
- The engine strips away HTML boilerplate, advertisements, and navigation elements.
- The resulting clean Markdown is returned for the agent to process or saved directly to a specified local directory.
Defuddle Web-to-Markdown Parser Setup
To integrate this tool into your Openclaw Skills workflow, ensure the CLI is installed globally on your system:
npm install -g defuddle
You can verify the installation by running defuddle --help in your terminal.
Defuddle Web-to-Markdown Parser Data Schema & Taxonomy
The skill supports various output structures depending on the user's requirements:
| Flag | Format Output |
|---|---|
--md |
Standard Markdown (recommended for LLMs) |
--json |
JSON object containing both HTML and Markdown |
-p <name> |
Specific metadata properties (title, description, domain) |
| No Flag | Returns the raw HTML content |
Defuddle Web-to-Markdown Parser Advanced Features
- Granular metadata extraction for specific page fields like title or domain.
- Direct file output management using the -o flag for automated document archival.
- Token-efficient parsing designed to fit more web content into limited AI context windows.
- Seamless integration with Obsidian-based note-taking workflows.
SKILL.md
Loading
Related Openclaw Skills
Native Run
A powerful execution bridge that allows OpenClaw agents to run native system commands on the local host machine.
Venice Transcribe
A powerful CLI tool for transcribing audio files to text using Venice AI high-accuracy Whisper models.
Venice Models
A CLI-powered discovery tool to explore, filter, and export Venice AI's comprehensive catalog of generative models and capabilities.
Venice Characters
An efficient CLI tool to browse, search, and integrate Venice AI's extensive library of character personas.
JSON Canvas Manager
A specialized skill for programmatically generating and managing .canvas files compatible with the JSON Canvas 1.0 specification.
Obsidian Bases
A technical skill for architecting .base files in Obsidian to create structured, database-like views of markdown note metadata.
Featured*
