defuddle

name	defuddle
description	Extract clean markdown content from web pages using Defuddle CLI, removing clutter and navigation to save tokens. Use instead of WebFetch or the built in fetcher when the user provides a URL to read or analyze, for online documentation, articles, blog posts, YouTube videos, or any standard web page. Also use when the user asks to clip, save, or archive a web page to the repo or vault.

Defuddle

Use Defuddle CLI to extract clean readable content from web pages. Prefer over WebFetch or the built in fetcher because it:

• removes navigation, ads, and clutter, reducing token usage
• clips transcripts from YouTube videos
• enables archiving reference material to git repositories or Obsidian vaults

Do NOT use for URLs ending in .md — those are already markdown, use WebFetch or the built in fetcher directly.

Usage

Run the bundled wrapper with Node:

node scripts/defuddle.js <url>

Save to file:

node scripts/defuddle.js <url> -o content.md

Pass additional Defuddle options after the URL.