Defuddle
My collection of agent skills
npx -y skills add jesse-black/skills --skill defuddleAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
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.
SKILL.md
1.0 KB, as published. Nobody here has run it
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.