Connect Puppeteer, Playwright or Selenium

Your automation client doesn't talk to Sendwin profiles directly — it connects over raw CDP (the Chrome DevTools Protocol). The flow is always the same: ask the local API to start a profile, take the WebSocket URL it returns, and point your client at that URL.

Because the profile launches with its real Sendwin Stealth Engine, fingerprint and proxy, everything your script does looks like the same distinct device the profile always uses.

Step 1 — start the profile

Call the start endpoint for the profile you want. It launches the browser and returns a CDP WebSocket URL in ws (also available as webSocketDebuggerUrl).

curl -X POST \
  -H "Authorization: Bearer YOUR_TOKEN" \
  http://127.0.0.1:36912/sw/profile/PROFILE_ID/start
# -> { "ok": true, "ws": "ws://127.0.0.1:.../devtools/browser/...", ... }

Step 2 — connect your client

Node + puppeteer-core:

const TOKEN = '...';
const start = await fetch(
  'http://127.0.0.1:36912/sw/profile/PROFILE_ID/start',
  { method: 'POST', headers: { Authorization: 'Bearer ' + TOKEN } }
).then(r => r.json());

const puppeteer = require('puppeteer-core');
const browser = await puppeteer.connect({ browserWSEndpoint: start.ws });
// ...drive the profile as usual

Playwright and Selenium follow the same idea: get the ws from the start call, then use your library's CDP-connect mechanism (for example Playwright's connectOverCDP, or a Selenium client configured to attach over CDP) with that endpoint.

If the profile is already running

You don't have to start it twice. If a profile is already open, fetch its current WebSocket instead:

curl -H "Authorization: Bearer YOUR_TOKEN" \
  http://127.0.0.1:36912/sw/profile/PROFILE_ID/ws
# -> { "ok": true, "ws": "..." }