Playwright MCP Server
Enables browser automation and web scraping with multi-session management, supporting page navigation, element interaction, network request capture, and content extraction across multiple concurrent browser instances.
README
Playwright MCP Server
基于 Playwright 的 Model Context Protocol (MCP) 服务器,提供浏览器自动化功能。
- 会话管理:创建和管理多个独立的浏览器会话
- 并发支持:支持多个并发浏览器会话,自动清理过期会话
- 浏览器操作:导航、点击、输入等常用操作
- 灵活配置:支持不同浏览器(Chromium、Firefox、WebKit)和模式(有头/无头)
- 自动清理:会话超时后自动清理资源
npm 地址:https://www.npmjs.com/package/@leviathangk/playwright-mcp
GitHub 地址:https://github.com/Leviathangk/PlaywrightMCPForCrawler
安装
npm i @leviathangk/playwright-mcp
MCP 配置
在 MCP Server 中添加:
{
"mcpServers": {
"playwright": {
"command": "node",
"args": [
"@leviathangk/playwright-mcp@latest",
"--browser", "chromium",
"--headless", "false",
"--session-timeout", "300000",
"--max-sessions", "10"
]
}
}
}
配置参数
| 参数 | 默认值 | 说明 |
|---|---|---|
--browser |
chromium |
浏览器类型:chromium、firefox、webkit |
--headless |
false |
无头模式 |
--session-timeout |
300000 |
会话超时(毫秒) |
--max-sessions |
10 |
最大并发会话数 |
--executable-path |
无 | 浏览器可执行路径(可选) |
核心工具
浏览器安装
browser_install- 安装 Playwright 浏览器(chromium、firefox、webkit)
会话管理
browser_create_session- 创建会话browser_close_session- 关闭会话
页面管理(多标签页)
browser_get_pages- 获取所有打开的页面列表browser_new_page- 创建新页面(标签页)browser_switch_page- 切换到指定页面browser_close_page- 关闭指定页面
页面操作
browser_navigate- 导航到 URLbrowser_click- 点击元素browser_type- 输入文本browser_scroll- 滚动页面browser_wait_for_element- 等待元素出现browser_execute_script- 执行自定义 JavaScript
页面分析
browser_get_page_structure- 获取页面结构browser_find_element_by_text- 根据文本查找元素browser_query_selector- 使用 CSS 选择器查询元素browser_get_page_content- 获取页面内容(HTML/文本/Markdown)browser_get_text_content- 获取元素文本内容browser_screenshot- 截图
网络请求捕获
browser_search_requests- 搜索请求(支持正则)browser_get_requests- 获取所有请求browser_get_request_detail- 获取请求详情(含 curl)browser_clear_requests- 清空请求历史
使用示例
基础使用
// 1. 创建会话
const session = await callTool('browser_create_session');
// 2. 导航到页面
await callTool('browser_navigate', {
sessionId: session.sessionId,
url: 'https://example.com'
});
// 3. 搜索 API 请求
const results = await callTool('browser_search_requests', {
sessionId: session.sessionId,
keyword: 'api/data',
searchIn: ['url']
});
// 4. 获取请求详情
const detail = await callTool('browser_get_request_detail', {
sessionId: session.sessionId,
requestId: results.matches[0].id
});
// 5. 关闭会话
await callTool('browser_close_session', {
sessionId: session.sessionId
});
多标签页管理
// 获取所有页面
const pages = await callTool('browser_get_pages', {
sessionId: sessionId
});
console.log(`总共 ${pages.totalPages} 个页面`);
pages.pages.forEach(page => {
console.log(`${page.index}: ${page.title} ${page.isActive ? '(当前)' : ''}`);
});
// 创建新页面
await callTool('browser_new_page', {
sessionId: sessionId,
url: 'https://www.example.com'
});
// 切换到指定页面
await callTool('browser_switch_page', {
sessionId: sessionId,
pageIndex: 1
});
// 关闭页面
await callTool('browser_close_page', {
sessionId: sessionId,
pageIndex: 2
});
页面分析
// 获取页面结构
const structure = await callTool('browser_get_page_structure', {
sessionId: sessionId,
maxElements: 50
});
// 查找元素
const element = await callTool('browser_find_element_by_text', {
sessionId: sessionId,
text: '登录'
});
// 点击元素
await callTool('browser_click', {
sessionId: sessionId,
selector: element.selector
});
// 获取页面内容
const content = await callTool('browser_get_page_content', {
sessionId: sessionId,
format: 'html'
});
// 执行自定义脚本
const result = await callTool('browser_execute_script', {
sessionId: sessionId,
script: 'return document.title;'
});
常见浏览器路径
Windows:
- Chrome:
C:\Program Files\Google\Chrome\Application\chrome.exe - Edge:
C:\Program Files (x86)\Microsoft\Edge\Application\msedge.exe
macOS:
- Chrome:
/Applications/Google Chrome.app/Contents/MacOS/Google Chrome - Edge:
/Applications/Microsoft Edge.app/Contents/MacOS/Microsoft Edge
Linux:
- Chrome:
/usr/bin/google-chrome - Chromium:
/usr/bin/chromium-browser
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.