mcp-browser-parallel
Multi-instance parallel browser MCP server that manages multiple isolated browser instances in a single Chrome process with automatic login state cloning, allowing parallel independent operations across instances.
README
mcp-browser-parallel
多实例并行浏览器 MCP Server — 在单个 Chrome 进程中管理多个隔离的浏览器实例,支持 Cookie 自动克隆。
核心特性
- 多实例隔离:每个实例拥有独立的 BrowserContext(cookies、storage、缓存完全隔离)
- 零登录克隆:从用户已登录的 Chrome 自动提取 auth,新实例自动继承登录状态
- 并行友好:所有操作通过
instanceId路由,无共享全局状态,多 AI 会话可同时操作不同实例 - CDP 窗口管理:支持通过 CDP 协议真正最大化窗口
工作原理
用户的 Chrome(已登录,端口 9222)
│
│ connectOverCDP
↓
┌─────────────────────────────────┐
│ mcp-browser-parallel Server │
│ │
│ browser_connect │
│ → 连接 Chrome │
│ → 提取 storageState │
│ │
│ instance_create("batch-1") │
│ → 新建 BrowserContext │
│ → 注入 cookies → 已登录! │
│ → 导航到目标 URL │
│ │
│ instance_create("batch-2") │
│ → 又一个独立 Context │
│ → 同样已登录! │
│ │
│ page_snapshot("batch-1") │ ← 操作 batch-1
│ page_click("batch-1", "e5") │ ← 不影响 batch-2
│ │
│ page_snapshot("batch-2") │ ← 操作 batch-2
│ page_click("batch-2", "e3") │ ← 不影响 batch-1
└─────────────────────────────────┘
安装
方式一:直接使用(推荐)
无需克隆代码,直接通过 npx 使用:
{
"mcpServers": {
"browser-parallel": {
"command": "npx",
"args": ["mcp-browser-parallel@latest"]
}
}
}
方式二:指定 Chrome CDP 端点
{
"mcpServers": {
"browser-parallel": {
"command": "npx",
"args": [
"mcp-browser-parallel@latest",
"--cdp-endpoint", "http://localhost:9222"
]
}
}
}
方式三:从源码构建
git clone <repo-url>
cd mcp-browser-parallel
npm install
npm run build
使用
1. 启动 Chrome(开启远程调试)
# Windows
chrome.exe --remote-debugging-port=9222
# macOS
/Applications/Google\ Chrome.app/Contents/MacOS/Google\ Chrome --remote-debugging-port=9222
# Linux
google-chrome --remote-debugging-port=9222
2. 在 Chrome 中登录目标系统
手动在 Chrome 中打开目标系统并完成登录。
3. 使用工具
Step 1: 连接 Chrome 并提取 auth
→ browser_connect(cdpUrl="http://localhost:9222")
Step 2: 创建隔离实例(自动带登录状态)
→ instance_create(instanceId="batch-1", url="https://your-system.com/module1")
→ instance_create(instanceId="batch-2", url="https://your-system.com/module2")
Step 3: 对每个实例独立操作
→ page_snapshot(instanceId="batch-1")
→ page_click(instanceId="batch-1", ref="e5")
→ page_snapshot(instanceId="batch-2")
→ page_click(instanceId="batch-2", ref="e3")
Step 4: 完成后关闭
→ instance_close_all()
工具清单
实例管理
| 工具 | 说明 |
|---|---|
browser_connect |
连接 Chrome CDP + 提取 auth |
instance_create |
创建隔离实例(自动克隆 auth) |
instance_list |
列出所有活跃实例 |
instance_close |
关闭指定实例 |
instance_close_all |
关闭所有实例 |
页面操作(所有操作通过 instanceId 路由)
| 工具 | 说明 |
|---|---|
page_navigate |
导航到 URL |
page_snapshot |
获取可访问性快照(含 ref) |
page_click |
点击元素(by ref) |
page_fill |
填入文本(by ref) |
page_type |
逐字输入(触发 input 事件) |
page_select_option |
下拉选择 |
page_hover |
悬停 |
page_press_key |
按键 |
page_screenshot |
截图 |
page_wait |
等待文本/时间 |
page_evaluate |
执行 JS |
page_maximize |
CDP 窗口最大化 |
与现有 MCP 对比
| 特性 | Chrome DevTools MCP | Playwright MCP | 本 MCP |
|---|---|---|---|
| 多实例隔离 | ❌ 共享 select_page 状态 | ❌ 单实例 | ✅ 独立 Context |
| Auth 克隆 | ❌ | ❌ | ✅ storageState |
| 并行安全 | ❌ 会话间冲突 | ❌ 会话间冲突 | ✅ instanceId 路由 |
| 窗口最大化 | ❌ | ✅ browser_run_code | ✅ 内置 CDP |
| 快照 + Ref | ✅ uid | ✅ ref | ✅ ref |
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.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
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.