Extract structured data from any web page by evaluating JavaScript via Playwright MCP

Agents frequently need to extract structured data from web pages — story titles from Hacker News, product prices from a catalog, metadata from documentation. Instead of parsing HTML server-side, the Playwright MCP browser_evaluate tool runs arbitrary JavaScript in the page's context with full DOM access. Navigate to the page, then pass a function that queries the DOM (querySelectorAll, textContent, href, etc.) and returns a clean JSON structure. No external parsing library needed — the browser IS the parser.