A text-first headless browser MCP server that extracts DOM and OCR text from web pages, offering a cheaper alternative to screenshot-based browsing by defaulting to zero image tokens.
