kepano/defuddle

Health v2 · maintenance only

Adoption not part of the score

9166 stars · 397 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Defuddle is a TypeScript library, CLI, and hosted service that extracts the main content from web pages, removing clutter like comments, sidebars, headers, and footers, and returns clean HTML or Markdown. It was created for the Obsidian Web Clipper browser extension and is designed as a more forgiving alternative to Mozilla Readability.

Use cases

• convert web articles to markdown
• extract main content from a webpage
• clean up html by removing sidebars and comments
• clip web pages into obsidian
• extract article metadata like title and author
• replace mozilla readability for content extraction
• parse html files from the command line

When to choose

• you need readable article content as markdown or clean html
• you want richer metadata extraction including schema.org data
• you need a readability alternative that removes fewer uncertain elements
• you want a cli or hosted api for url-to-markdown conversion

When to avoid

• you need full-page screenshots or pixel-perfect rendering
• you require a battle-tested, fully stable parser - the project is a work in progress
• you need structured data scraping like tables or product listings rather than article content

Facets

library · maturity active

parser web-scraping cli markdown web-development developer-tools crawlers browser cli readability content-extraction html-to-markdown article-extraction obsidian web-clipper nodejs web-server

Member repositories

For agents

markdown · JSON · MCP: product_card(name="kepano/defuddle")