My existing Python/Selenium script used to capture full-page screenshots from several e-commerce sites, but the stores have tightened their defences and the script is now blocked. The main hurdles I’m seeing are IP blocking and user-agent detection; no CAPTCHA so far, but that could appear at scale. I’m not running any proxy rotation yet, so every request currently comes from a single IP and standard headless-browser signature—precisely what I need to change. The goal is a reproducible workflow that can visit a page, fully render dynamic content, and save a high-quality screenshot without tripping bot filters—then repeat the process reliably across a million URLs over time. Key points you should address inside the solution: • Work comfortably with e-commerce architectures that rely on script-heavy, client-side rendering. • Bypass IP blocking and basic fingerprint checks (user-agent, headless flags, navigator properties, etc.). • Handle concurrency and throttling so pages load quickly but don’t raise red flags. • Deliver cleanly named image files (PNG or JPEG) plus minimal meta-data (timestamp, resolved URL). • Keep the stack in Python unless a demonstrably stronger approach (e.g., Playwright, Puppeteer, or a stealth-browser fork) is justified and still callable from a Python driver. What I’d like to receive: 1. A production-ready script or small package (Docker is fine) that accepts a URL list and outputs screenshots. 2. Brief documentation so I can extend the setup to millions of URLs on cloud instances. If you can demo a short run that sails past the current blockers, we’re ready to move forward quickly.