jo-inc / camofox-browser
A C++-level anti-detection browser server that gives AI agents stealthy, token-efficient web access where Playwright gets blocked
星标趋势
AI 分析
项目摘要
Camofox-browser is a stealth headless browser server designed for AI agents, wrapping Camoufox (a Firefox fork with C++ level fingerprint spoofing) in a REST API. It provides accessibility snapshots, stable element references, and anti-detection capabilities to bypass Cloudflare, bot detection, and anti-scraping measures that block standard Playwright/Puppeteer browsers.
为什么值得关注
It addresses a critical pain point for AI agents operating on the real web — modern bot detection increasingly blocks standard headless browsers. By patching Firefox at the C++ level rather than relying on JavaScript shims, it achieves more robust fingerprint evasion. The token-efficient accessibility snapshot approach is also a thoughtful design choice for LLM-based agents that pay per token.
优势
- C++ level fingerprint spoofing is more robust than JavaScript-based stealth plugins
- Token-efficient accessibility snapshots reduce LLM API costs by ~90% vs raw HTML
- Lightweight design (~40MB idle) enables deployment on low-resource infrastructure like Raspberry Pi or $5 VPS
- Active development with 14 releases in 6 months, CI, tests, and Docker support
局限性
- 115 open issues suggests the project is still maturing and may have stability gaps
- No examples directory, which could slow onboarding for new users
- Relies on a specific Firefox fork (Camoufox) which may lag behind upstream Firefox security patches
使用场景
- AI agent web browsing where standard headless browsers get blocked by Cloudflare or bot detection
- Web scraping at scale with reduced detection risk
- Automated testing of web applications from an agent perspective
- Cost-efficient web interaction for LLM-powered agents via compact accessibility snapshots