Bridges AI apps to the LionScraper browser extension for Chrome and Edge, letting models scrape lists, articles, links, images, and other structured data from live web pages. Ships as both npm and PyPI packages, each exposing an MCP stdio server plus a CLI daemon that also serves an HTTP API on localhost:13808. The extension does the actual scraping with full JavaScript execution. Without the extension connected, it falls back to basic HTTP fetch mode. Useful when you need your AI to pull real page content that requires a browser runtime, not just raw HTML. Port numbers must match between your MCP config and the extension settings, and you'll want to keep target pages open during scraping.
claude mcp add --transport stdio io.github.dowant-lionscraper-node uvx lionscraper-node