agent-stealth-browser
A stealth-enhanced browser automation MCP server for AI agents to interact with websites while bypassing anti-bot detection mechanisms like Cloudflare and reCAPTCHA.
README
Agent Stealth Browser
🕵️ 通用的 AI 浏览器自动化 MCP 工具,让 AI Agent 能够以 Stealth 模式 智能地操作任意网站。
Agent Stealth Browser 是 agent-browser 的 Stealth 增强版本,提供完全兼容的 API 接口,但具备绕过反爬检测的能力。
✨ 特性
- 🔄 完全兼容 agent-browser API - 无缝替换,相同的工具名和参数格式
- 🕵️ Stealth 模式 - 绕过 Cloudflare、reCAPTCHA 等反爬检测
- 🌍 跨平台支持 - Windows、macOS、Linux 自动适配
- 🍪 Cookie 持久化 - 保存登录状态,跨会话使用
- 📸 截图能力 - 支持全页面或指定元素截图
- 🎯 @ref 元素定位 - 与 agent-browser 一致的元素引用格式
📦 安装
方式 1: uvx (推荐)
uvx agent-stealth-browser
方式 2: pip
pip install agent-stealth-browser
# 安装 Playwright 浏览器
playwright install chromium
# 可选: 安装 stealth 增强
pip install playwright-stealth
方式 3: npx
npx agent-stealth-browser
🔧 MCP 配置
Claude Desktop / Cursor
在 MCP 配置文件中添加:
{
"mcpServers": {
"stealth-browser": {
"command": "uvx",
"args": ["agent-stealth-browser"]
}
}
}
或使用 npx:
{
"mcpServers": {
"stealth-browser": {
"command": "npx",
"args": ["agent-stealth-browser"]
}
}
}
🛠️ 可用工具
| 工具名 | 描述 | 对应 agent-browser 命令 |
|---|---|---|
browser_navigate |
导航到指定 URL | open <url> |
browser_snapshot |
获取页面交互元素快照 | snapshot -i |
browser_click |
点击元素 | click @e1 |
browser_fill |
填写输入框(先清空) | fill @e1 "text" |
browser_type |
逐字输入(模拟真人) | type @e1 "text" |
browser_press |
按下键盘按键 | press Enter |
browser_scroll |
滚动页面 | scroll down 500 |
browser_hover |
悬停在元素上 | hover @e1 |
browser_select |
选择下拉框选项 | select @e1 "value" |
browser_get |
获取元素/页面信息 | get text @e1 |
browser_eval |
执行 JavaScript | eval "code" |
browser_screenshot |
页面截图 | screenshot |
browser_wait |
等待时间或元素 | wait |
browser_cookies_save |
保存 cookies | - |
browser_cookies_load |
加载 cookies | - |
browser_close |
关闭浏览器 | close |
📝 使用示例
基础浏览
1. browser_navigate: {"url": "https://example.com"}
2. browser_snapshot: {}
3. browser_click: {"target": "@e1"}
表单填写
1. browser_navigate: {"url": "https://example.com/login"}
2. browser_snapshot: {}
3. browser_fill: {"target": "@e3", "text": "username"}
4. browser_fill: {"target": "@e4", "text": "password"}
5. browser_click: {"target": "@e5"}
保持登录状态
# 登录后保存 cookies
browser_cookies_save: {"domain": "example"}
# 下次启动时加载
browser_navigate: {"url": "https://example.com"}
browser_cookies_load: {"domain": "example"}
browser_navigate: {"url": "https://example.com"} # 刷新页面应用 cookies
🌍 跨平台支持
Agent Stealth Browser 自动检测运行平台并适配:
| 平台 | User-Agent | 浏览器参数 |
|---|---|---|
| macOS | Macintosh; Intel Mac OS X | 标准参数 |
| macOS (Apple Silicon) | Macintosh; Apple M1 | 标准参数 |
| Windows | Windows NT 10.0; Win64 | --disable-gpu |
| Linux | X11; Linux x86_64 | --no-sandbox 等 |
🔒 Stealth 模式
默认启用以下反检测措施:
- ✅ 隐藏
navigator.webdriver属性 - ✅ 模拟真实 Chrome 环境
- ✅ 隐藏自动化特征
- ✅ 支持
playwright-stealth插件(可选增强)
增强 Stealth 模式
安装 playwright-stealth 获得更强的反检测能力:
pip install playwright-stealth
📁 数据存储
- Cookies:
~/.agent_stealth_browser/cookies/
🔄 与 agent-browser 的区别
| 特性 | agent-browser | agent-stealth-browser |
|---|---|---|
| Stealth 模式 | ❌ | ✅ |
| 反爬绕过 | ❌ | ✅ |
| Cookie 持久化 | ❌ | ✅ |
| 跨平台适配 | ✅ | ✅ |
| API 兼容性 | - | 100% 兼容 |
🤝 贡献
欢迎提交 Issue 和 Pull Request!
📄 License
MIT License - 详见 LICENSE 文件
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.