defuddle

name	defuddle
description	Extract clean markdown from web pages using the Defuddle CLI — removes navigation, ads, and clutter to save tokens. Use instead of WebFetch when the user provides a URL to read or analyze (articles, docs, blog posts). Do NOT use for URLs ending in .md (already markdown — use WebFetch directly). Requires: npm install -g defuddle
model	haiku
allowed-tools	Bash

Defuddle

Extracts clean, token-efficient markdown from web pages. Preferred over WebFetch for article/documentation URLs.

Basic Usage

# Extract to stdout (markdown)
defuddle parse <url> --md

# Save to file
defuddle parse <url> --md -o content.md

# Get metadata only
defuddle parse <url> -p title
defuddle parse <url> -p description
defuddle parse <url> -p author

When to Use

• Articles, blog posts, documentation pages
• Any page where you want content without navigation/ads

When NOT to Use

• URLs ending in .md → use WebFetch directly
• JS-heavy SPAs that require execution → use Playwright fallback
• Anti-bot protected pages → use Camoufox MCP

Install (once)

npm install -g defuddle