What browsers are supported?
Chromium-based browsers primarily. Chrome, Chromium, Edge (Chromium), and Brave are supported. Firefox requires experimental configuration.
How do I handle login-required sites?
Configure session persistence through cookie storage. For SSO sites, automate credential entry or use pre-authenticated sessions.
What's the maximum page size I can extract?
Default limit is 10MB of content. Larger pages are chunked for MCP transmission. Configure MAX_CONTENT_SIZE_MB for adjustments.
Can I run in non-headless mode?
Yes, set BROWSER_HEADLESS=false. Useful for debugging but not recommended for production automation.
How do I prevent being blocked by sites?
Configure appropriate user agents, respect robots.txt, implement rate limiting, and use proxy rotation for high-volume scraping.
What happens if a browser crashes?
The server detects crashes and spawns fresh instances. Operations fail gracefully with error messages for retry.
Do I need special Chrome installation?
Standard Chrome installation works. For servers, use headless-compatible builds or Puppeteer's bundled Chromium.
How do I handle CAPTCHA?
CAPTCHA solving requires third-party services. Configure CAPTCHA_API_KEY for integrated solving. Some sites may require manual intervention.
Can I interact with iframes?
Yes, the server handles iframe context switching automatically. Element selectors work across frame boundaries.
What about infinite scroll pages?
Configure MAX_SCROLLS and SCROLL_PAUSE_MS for infinite scroll handling. The server scrolls and collects content incrementally.
Who should use this?
Use this guide if you are shortlisting AI agents for a real workflow in Bengaluru, Mumbai, Delhi NCR, Hyderabad, Kochi, Pune, Chennai, or a distributed Indian team serving global clients. The strongest use cases include coding productivity, WhatsApp automation, CRM updates, customer support, appointment booking, lead qualification, knowledge-base retrieval, workflow automation, and internal reporting.
Who should avoid this?
Avoid treating this page as final procurement approval if your workflow involves regulated financial, medical, legal, HR, or child-related personal data. In those cases, use the page as a shortlist and then run vendor security review, DPDP impact review, access-control review, and a controlled pilot with human oversight.