cdpBaseUrl, and pass it directly to playwright.connectOverCDP() or Puppeteer’s connect(). From there, every Playwright or Puppeteer API you already know works exactly as it does locally — navigate, click, fill forms, intercept network requests, and extract data — while AdsCrawl handles the browser infrastructure, proxies, and fingerprinting.
How it works
Create a session
Send POST /cdp/sessions with your desired idleTimeoutMs, maxSessionMs, and browserSettings. The API responds with a sessionId and a cdpBaseUrl that already contains your data token.
Connect with Playwright or Puppeteer
Pass cdpBaseUrl directly to your automation library. No additional configuration is needed.
Automate the page
Navigate, interact, and extract data using the full Playwright or Puppeteer API surface — the remote browser behaves identically to a local one.
Delete the session when done
Call DELETE /cdp/sessions/:sessionId to terminate the browser and release resources immediately, rather than waiting for the idle or max-session timeout to expire.
Security model
Remote CDP uses two separate tokens to protect your session:- Data token — embedded in
cdpBaseUrland used for all CDP WebSocket connections and discovery requests. This token is long-lived (valid for the session lifetime) and scoped to read/control the browser over CDP. - Control token (
controlToken) — a single-use token issued byPOST /cdp/live-tokenfor live interactive viewer access. It expires after 30 seconds and can only be consumed once. Request a fresh control token each time you need to open the live viewer.
Session lifecycle
Configure the session lifetime at creation time with two fields:GET /cdp/sessions at any time, or delete a specific session with DELETE /cdp/sessions/:sessionId.