网络爬虫隐身技术栈选型与Bot检测规避方案生成器

You are a senior web scraping and anti-detection engineer. I need you to design a comprehensive **Stealth Crawling Architecture** for an AI agent that needs to reliably access web data without being blocked. ## Requirements - Target sites: [e.g., E-commerce / Social media / SaaS platforms] - Scale: [e.g., 10K pages/day / 1M pages/day] - Data freshness: [e.g., Real-time / Daily / Weekly] - Budget: [e.g., $0 (open source only) / $100-500/mo / Enterprise] ## Please Provide: ### 1. Browser Engine Selection Compare and recommend from: - Stealth Chromium forks (CloakBrowser, Camoufox, etc.) - Playwright with stealth plugins - Puppeteer-extra-stealth - Headless detection bypass patches For each, evaluate: detection pass rate, maintenance status, resource usage, ease of integration. ### 2. Fingerprint Randomization Strategy - Canvas/WebGL fingerprint spoofing - Navigator/UA rotation with consistency rules - Timezone/locale/language coherence - Screen resolution and device memory patterns - WebRTC leak prevention ### 3. Network Layer - Proxy pool architecture (residential vs datacenter vs mobile) - IP rotation cadence per target - TLS fingerprint (JA3/JA4) randomization - DNS-over-HTTPS configuration ### 4. Behavioral Mimicry - Mouse movement patterns (Bezier curves, jitter) - Scroll behavior simulation - Typing cadence for form fills - Session duration and page dwell time distributions ### 5. Detection Test Checklist Provide a pass/fail checklist against: - Cloudflare Turnstile - DataDome - PerimeterX/HUMAN - Akamai Bot Manager - reCAPTCHA v3 score ### 6. Architecture Diagram Output a deployment architecture with Docker Compose showing proxy rotation, browser pool, task queue, and result storage. Format: Structured markdown with comparison tables and a Mermaid architecture diagram.