Automation & Data Scraping

Why scrapers need isolated environments

Blocks are usually not about the IP alone — it's the combination of fingerprint + behavior + IP. Giving each scraping session its own fingerprint and proxy treats "a real device" as the unit of scraping:

Scraping pain	Traditional fix	Virtual Browser fix
Detected as headless	UA spoofing with patchwork	Real Chromium core + complete fingerprint
Same-IP blocks	Rotating proxy pool	Proxy pool × fingerprint pool — double dispersion
Lost login state	Re-login triggers risk control	Environments keep cookies; sessions are reusable
Complex orchestration	Self-managed browser clusters	REST API / CLI scheduling of environments

Three integration paths

1. Local REST API (Pro)

Environments as resources, CRUD-style scheduling:

# Create and start a proxy-bound environment
curl -X POST http://127.0.0.1:PORT/api/env/create \
  -H 'Content-Type: application/json' \
  -d '{"proxy": "socks5://user:pass@host:port"}'

# Response includes the CDP endpoint for Playwright / Puppeteer

2. Playwright over CDP

const { chromium } = require('playwright')

// cdpUrl is returned after the environment starts
const browser = await chromium.connectOverCDP(cdpUrl)
const page = await browser.contexts()[0].pages()[0]

await page.goto('https://example.com', { waitUntil: 'domcontentloaded' })
const data = await page.$$eval('.item', els =>
  els.map(el => el.textContent.trim())
)
// Close when done — fingerprint and cookies persist in the environment

3. MCP: let an AI agent drive it

Virtual Browser ships an MCP service, so AI tools like Cursor can call "create environment / start / open page" directly and turn scraping flows into natural-language tasks.

User: Scrape prices from these 20 product pages, one environment each
AI: Started 20 isolated environments (unique fingerprints + rotating proxies), scraping done, results in CSV.

Pacing suggestions

Target difficulty	Suggested setup
Lenient (docs, public data)	A few environments, low rotation
Medium (e-commerce listings, search)	Environment pool + residential proxies + random dwell
Strict (social media, logged-in content)	Environments bound to accounts long-term, human-like pacing

Session reuse is an underrated lever

Environments keep cookies and local storage, so after one login you can visit at low frequency for a long time — far more human-like (and cheaper) than "fresh environment + re-login every time".

FAQ

Q: Is the API available on the free plan? GUI management of isolated environments is free forever; REST API, window synchronizer, MCP, and cloud sync are Pro features. Prototype on free, then upgrade if needed.

Q: How is this different from plain Playwright + proxies? Playwright's own browser fingerprint is public and uniform. Every Virtual Browser environment is a complete, stable, reusable fingerprint identity — and the GUI makes manual intervention easy.

Q: Can scraped data feed my pipeline directly? Yes. Environments are just the browser layer; extraction and storage stay in your Playwright / Puppeteer code, fully under your control.

Next: read the automation API guide and the API reference, or download the client to start.