ScrapeNinja gives Claude two endpoints for web scraping: a fast non-JS mode with Chrome TLS fingerprinting, and a full browser mode for JavaScript-heavy sites. The smart retry system lets you specify status codes or text patterns that should trigger another attempt, which is handy for flaky targets. It includes geo-based proxies across US, EU, and a few other regions, plus a Cheerio extractor for pulling structured data without writing separate parsing code. The AJAX interception feature is genuinely useful if you're trying to grab API responses that populate a page. Start with the fast endpoint and only move to JS rendering if you need it, since it's slower and burns more credits.
npx skills add https://github.com/vm0-ai/vm0-skills --skill scrapeninja