Web scraping and browser automation have always been an arms race. As sites like LinkedIn, Cloudflare, and FingerprintJS deploy increasingly sophisticated detection mechanisms, traditional tools like Playwright and Selenium find themselves blocked before they can even load a page. CloakBrowser is a drop-in Playwright replacement built on a heavily modified Chromium that passes virtually every anti-bot check — and it's gaining serious traction in the AI agent community.

• Native Fingerprint Randomization at the Binary Level: CloakBrowser doesn't just mask JavaScript properties — it modifies Chromium at the C++ level to generate authentic hardware fingerprints. This includes canvas rendering, WebGL vendor/renderer strings, audio context signatures, and font lists. The result is a browser that looks indistinguishable from a real user browser to fingerprinting scripts.
• Closed Shadow DOM Support: Using Chrome's --enable-blink-features=FakeShadowRoot flag, CloakBrowser exposes closed shadow roots that would normally be inaccessible, resolving one of the most common anti-bot workarounds used by Cloudflare Turnstile and similar services.
• Drop-in Playwright API Compatibility: If you already use Playwright, switching to CloakBrowser requires just a one-line import change. The project provides cloakbrowser as a Python package with full API parity, including launch(), launchPersistentContext(), and all the familiar locator/expect patterns.
• Mobile Device Emulation: Full mobile fingerprint support including viewport dimensions, device scale factor, touch capabilities, and user agent strings — all passed through at the binary level rather than through JavaScript emulation.

Beyond the README, the project's GitHub Issues reveal a vibrant community actively working through real-world detection challenges:

Users reported mixed results against BrowserScan's fingerprint authenticity scoring. The maintainer clarified an important distinction:

"BrowserScan bot detection passes clean on our end (17/17 Normal). The flags you're seeing are on the fingerprint authenticity page, which is a different check. We've investigated BrowserScan fingerprint scoring extensively (#31, #49). With the right config you can get a clean score — key settings are headless=False and geoip=matching."
— Cloak-HQ (maintainer)

A developer shared detailed screenshots of FingerprintJS API responses showing bot flags:

"Can you send a screenshot of the page that shows they detected you? And also send a screenshot of what you see when you visit: https://demo.fingerprint.com/playground — we want to compare our baseline against your specific setup."
— eofreternal

The thread evolved into a deep technical discussion about WebGL renderer strings and GPU memory signatures, with the maintainer promising a dedicated configuration guide for FingerprintJS evasion.

A critical bug where authenticated SOCKS5 proxies would silently bypass authentication:

"Build `.2` (146.0.7680.177.2) added native SOCKS5 credential handling at the binary level. Your environment section lists `.1` though — can you confirm which binary you actually have? Run npx cloakbrowser info to check."
— Cloak-HQ (maintainer)

CloakBrowser represents a significant step forward in the browser automation arms race. By operating at the Chromium binary level rather than patching JavaScript APIs, it achieves a level of fingerprint authenticity that previous tools simply couldn't match. With an active maintainer responding to issues within hours and a community that's collectively stress-testing against dozens of detection services, it's one of the most compelling open-source projects in the AI agents and web scraping space right now. Whether you're building a browser-use agent, automating data collection at scale, or just tired of seeing hCaptcha blocks, CloakBrowser is worth a serious look.

⭐ 4,717 GitHub Stars · Python · MIT License
🔗 https://github.com/CloakHQ/CloakBrowser · Maintained by @Cloak-HQ