When you need to analyze web content without burning tokens on navigation menus and ads, this extracts clean markdown from any URL. It's essentially a smart web scraper that strips the clutter and gives you just the article content. Perfect for feeding documentation, blog posts, or news articles into Claude for analysis or summarization. The CLI tool handles the heavy lifting of content detection and cleanup, so you get readable markdown instead of messy HTML soup. Skip it for URLs that already end in .md since those don't need the extraction treatment.
npx skills add https://github.com/kepano/obsidian-skills --skill defuddle