ask-html-renderer-api

POST to /api/render with the target url and your SCRAPER_API_KEY as a Bearer token. Set format to "json" for structured output with title, html, and cookies, and tune waitUntil, timeout, and blockImages for load behavior.

POST credentials to /api/login with a sessionId, then call /api/fetch with the same sessionId. Cookies are saved into Durable Object storage after each render and reloaded on the next, so the session persists across requests.

The key must exactly match the SCRAPER_API_KEY binding, and authentication runs before routing on every route including /api/login and /api/fetch. The key can be sent as a Bearer header, query parameter, or POST body field.

Yes, set bypassCaptcha to true to attempt Cloudflare-challenge bypass, and provide a twoCaptchaKey for reCAPTCHA or Turnstile. Tune challengeMatch, maxRetries, and challengeTimeout when the interstitial markup is unusual.

Cookie persistence is scoped per sessionId, with one Durable Object instance and cookie jar per id. If you change or omit the sessionId, you get the "default" jar, so reuse a consistent sessionId to keep sessions alive.

The package.json start script points at a .js file while the source is TypeScript, and the project is a Cloudflare Worker, not a Node app. Use wrangler dev or wrangler deploy with a wrangler config binding MYBROWSER, BROWSER_DO, and SCRAPER_API_KEY.