Collect the exact numbers you need from website pages, even when sign in is required. Great for teams that track competitor stats, pricing, or profile metrics and want simple, reliable results on demand.

A webhook receives a JSON request with a subject or a direct page link plus up to five target data points. If only a subject and domain are sent, the flow searches and picks the right URL. A Selenium browser session starts, applies anti detection settings, and can use a proxy. Cookies can be injected to reach gated pages. The browser loads the page, resizes the window, and takes screenshots. Files are then passed to an OpenAI model that reads the page image and returns the requested fields. Clear responses report success or common errors, and all sessions are closed safely.

Setup needs a Selenium container, an OpenAI API key, and a proxy if you plan to scrape at scale. You may also capture session cookies with a helper extension to access private areas. Expect large time savings by turning long manual checks into quick API calls. Ideal uses include checking public follower counts, reading product prices, and pulling values from dashboards that require login.

• Webhook intake that accepts subject, domain or direct URL, target data list, and cookies.
• Search and selection step that finds the best matching page and extracts the first valid link.
• Selenium session creation with anti detection script to mask automation signals.
• Optional cookie injection to access pages after login without manual steps.
• Browser control for window resize, refresh, and navigation stability.
• Screenshot capture and file conversion for reliable page snapshots.
• OpenAI GPT 4o analysis of screenshots to extract the requested fields.
• Proxy friendly design with an IP check branch to confirm routing is correct.
• Robust error responses and session cleanup to keep the system stable.

• Reduce manual website checks from hours to minutes by sending a single webhook request.
• Automate up to 5 target metrics per call to cover the most important numbers fast.
• Reach gated pages with cookie based sessions and cut manual logins to zero.
• Lower blocks and retries by routing traffic through a residential proxy.
• Improve accuracy by using image based extraction and reduce copy paste errors by 80%.
• Handle more pages per day with a stateless webhook that can scale as needed.

1. Import the template into n8n: Create a new workflow in n8n > Click the three dots menu > Select 'Import from File' > Choose the downloaded JSON file.
2. You'll need accounts with OpenAI. See the Tools Required section above for links to create accounts with these services.
3. In the n8n credentials manager, create an OpenAI credential: open Credentials > New > OpenAI API > paste your API key from your OpenAI account page, give the credential a clear name, and save.
4. Deploy a Selenium Chrome container and ensure it is reachable at http://selenium_chrome:4444/wd/hub. Match this host and port with the HTTP Request nodes that control Selenium.
5. If using a proxy, whitelist your server IP with your proxy provider. In the Create Selenium Session node, add the argument --proxy-server=address:port. Selenium does not support proxy authentication, so IP whitelisting is required.
6. If you need login access, capture session cookies with your preferred browser extension and keep them as a JSON array. You will pass this array in the webhook request.
7. Open the Webhook node and copy the Test URL. Use this URL for initial tests, then switch to the Production URL when ready.
8. Review the Code and Inject Cookie nodes and confirm they reference the cookies field from the incoming request. No change is needed if you follow the sample JSON structure.
9. Check the Limit node to keep the Target data list to five items. Sending more than five will be blocked by design.
10. Send a test request with a direct Target Url to skip search. Then test with subject and domain to confirm the search and link extraction path is working.
11. Inspect execution data: verify screenshots in the Convert to File nodes and review the JSON response. If you see a blocked message, add a residential proxy or adjust timing with the refresh step.
12. When stable, move to the Production Webhook URL, restrict access by IP or token if needed, and monitor session cleanup in the Delete Session nodes.