Scraper

Gives Claude five tools to pull content from web pages without needing Firecrawl or an API key. Uses Mozilla Readability under the hood, the same engine that powers Firefox Reader View. You get scrape_url for clean markdown extraction, extract_links for all hrefs and anchor text, extract_metadata for OG tags and page info, search_page to find text matches, and scrape_multiple for batch jobs. Works well on server-rendered docs and blogs. Won't handle heavy JavaScript SPAs, but if you're asking Claude to summarize a docs page or compare API references across sites, this does it without manual copy-paste.