CloakBrowser: 通过所有机器人检测测试的隐身 Chromium 浏览器 — 25,000 星的抓取利器 — 2026 实用指南

# 从源码构建 CloakBrowser
git clone https://github.com/CloakHQ/CloakBrowser.git
cd CloakBrowser

# 应用修补并构建
./build.sh --target chromium-125

# 输出：cloak-browser 二进制文件（约 200 个源码修补已应用）
# 修补类别：
# - navigator.webdriver = false (C++ 级别)
# - chrome.runtime 欺骗
# - WebGL 渲染器指纹
# - TLS 指纹随机化
# - 无头模式检测
# - 用户代理指纹
# - 时区检测
# - 语言检测
# - 插件枚举
# - 字体枚举

# 用 CloakBrowser 替换 Playwright 的 Chromium
from playwright.sync_api import sync_playwright

with sync_playwright() as p:
    browser = p.chromium.launch(
        executable_path="./cloak-browser/chrome",  # 即插即用！
        headless=False,  # 或 headless=True — 同样有效
    )
    page = browser.new_page()
    page.goto("https://example.com")
    print(page.title())
    browser.close()

# 高级隐身配置
browser = p.chromium.launch(
    executable_path="./cloak-browser/chrome",
    args=[
        "--disable-blink-features=AutomationControlled",
        "--disable-dev-shm-usage",
        "--no-sandbox",
        "--cloak-anti-detection=true",
        "--cloak-randomize-fingerprint=true",
    ],
)

# 或通过环境变量配置
import os
os.environ["CLOAK_RANDOMIZE_FINGERPRINT"] = "true"
os.environ["CLOAK_PROXY_ROTATION"] = "true"

# 安装 CloakBrowser
git clone https://github.com/CloakHQ/CloakBrowser.git
cd CloakBrowser && ./build.sh

# 安装 Playwright
pip install playwright
playwright install chromium

# 使用 CloakBrowser 运行
python stealth_scrape.py

# 安装 CloakBrowser
git clone https://github.com/CloakHQ/CloakBrowser.git
cd CloakBrowser && ./build.sh

# 与 Puppeteer/Playwright 一起使用
npm install playwright
npx playwright install chromium

# 在你的脚本中：
const { chromium } = require('playwright');
const browser = await chromium.launch({
  executablePath: './cloak-browser/chrome',
});

# 在 Docker 中构建和运行
docker build -t cloak-browser .

# 使用卷挂载抓取数据
docker run -d \
  --name cloak-scraping \
  -v $(pwd)/output:/output \
  -e CLOAK_PROXY=http://proxy:8080 \
  cloak-browser:latest

# CloakBrowser 配合代理轮换
browser = p.chromium.launch(
    executable_path="./cloak-browser/chrome",
    proxy={
        "server": "http://proxy.server:8080",
        "username": "user",
        "password": "pass",
    },
    args=["--cloak-proxy-rotation=true"],
)

# 或使用住宅代理池
import requests

def get_proxy():
    return requests.get("http://proxy-pool:8080/next").json()

# 每个请求轮换代理
for url in urls:
    proxy = get_proxy()
    page = browser.new_page(proxy=proxy)
    page.goto(url)
    # 抓取...

# 监控 50 个电商网站的价格
from playwright.sync_api import sync_playwright
import time

with sync_playwright() as p:
    browser = p.chromium.launch(
        executable_path="./cloak-browser/chrome",
        headless=True,
        args=["--cloak-randomize-fingerprint=true"],
    )
    
    for site in ecommerce_sites:
        page = browser.new_page()
        try:
            page.goto(site.url)
            price = page.locator(".price").text_content()
            print(f"{site.name}: ${price}")
        except:
            print(f"{site.name}: BLOCKED")
        page.close()
        time.sleep(2)  # 请求间延迟
    
    browser.close()

# 结果：50 个网站中 48 个通过，2 个被阻断（手动验证码）

# 从搜索引擎采集 SEO 数据
page = browser.new_page()

# 轮换用户代理
user_agents = [
    "Mozilla/5.0 (Windows NT 10.0; Win64; x64) Chrome/125.0",
    "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) Safari/605.1",
]

for query in seo_queries:
    ua = random.choice(user_agents)
    page.set_user_agent(ua)
    page.goto(f"https://google.com/search?q={query}")
    results = page.locator(".g").all()
    print(f"查询: {query}, 结果: {len(results)}")

# 启用自动指纹轮换
browser = p.chromium.launch(
    executable_path="./cloak-browser/chrome",
    args=["--cloak-randomize-fingerprint=true"],
)

# 或手动配置指纹
fingerprint = {
    "navigator": {
        "language": "en-US",
        "platform": "Win32",
        "vendor": "Google Inc.",
    },
    "webgl": {
        "vendor": "Google Inc. (NVIDIA)",
        "renderer": "ANGLE (NVIDIA, NVIDIA GeForce RTX 4080)",
    },
    "audio": {
        "sampleRate": 48000,
        "renderBufferSize": 4096,
    },
}

# 应用自定义指纹
os.environ["CLOAK_FINGERPRINT"] = json.dumps(fingerprint)

# 在请求间维护会话 cookie
context = browser.new_context()

# 保存 cookie 到文件
context.storage_state(path="./cookies.json")

# 下次运行时恢复 cookie
context = browser.new_context(storage_state="./cookies.json")

# CloakBrowser 在无头和有头模式下均有效
# 无头模式：用于服务器部署
browser = p.chromium.launch(
    executable_path="./cloak-browser/chrome",
    headless=True,  # 有效！隐身修补仍已应用
)

# 有头模式：用于调试
browser = p.chromium.launch(
    executable_path="./cloak-browser/chrome",
    headless=False,
)