Browser Use
Browser agents and stealth cloud browsers for AI: run web tasks, fill forms, extract data.
作者提供的说明(英文原文)
Web data for agents. Extract web pages, crawl websites, discover URLs, and search the web.
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
X-Api-Key标为机密的值是你账户的钥匙。千万不要分享给别人,也不要粘贴到聊天里。
你需要
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
这个地址需要在请求的 header 中放入密钥,而添加连接器的菜单无法设置。请改用 Claude Code、Cursor 或 VS Code。
在终端中运行这条命令
claude mcp add --transport http geonode https://mcp.geonode.com/mcp --header 'X-Api-Key: your-value'运行前请先填写空着的值。
粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"geonode": {
"url": "https://mcp.geonode.com/mcp",
"headers": {
"X-Api-Key": "your-value"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"geonode": {
"type": "http",
"url": "https://mcp.geonode.com/mcp",
"headers": {
"X-Api-Key": "your-value"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
这个地址需要在请求的 header 中放入密钥,而添加连接器的菜单无法设置。请改用 Claude Code、Cursor 或 VS Code。
你的选择通过 Cookie 保存在这台设备上,无需账户。
X-Api-Key 机密 HTTP header这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
我们还没有读取这个服务器的工具列表;我们会轮流检查各个端点。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月6日
注册名称已通过域名 geonode.com 验证(注册中心会核实域名所有权)
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
尚未检查;我们会轮流检查各个端点
尚未检查;我们会轮流检查各个端点
尚未检查;我们会在检查端点时读取工具列表
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
你需要填写:X-Api-Key
它在作者的云端运行。通过互联网连接,无需安装,但你发送的内容会经过作者的服务器。
GitHub 仓库最后更新于–,最新版本(1.0.1)于10天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接,所以接受远程 MCP 服务器的应用也可以使用,例如开启开发者模式的 ChatGPT。
同一分类中有:Browser Use, Chrome DevTools MCP, MCP Server Browser
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
Browser agents and stealth cloud browsers for AI: run web tasks, fill forms, extract data.
MCP server for Chrome DevTools
MCP server for browser use access
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
Give MCP-compatible AI assistants web search and content collection tools. Extract pages, discover URLs, crawl sites and run batches for research and downstream analysis. Geonode offers limited keyless page extraction; an API key enables the full toolset, including JavaScript rendering and country targeting.
[Try Geonode for free](https://app.geonode.com/sign-up)
Integrations page · Geonode API docs · MCP setup guide · View source on GitHub · Official MCP Registry
Connect · Example: extract a page and answer with sources
| Setting | Value |
|---|---|
| Remote endpoint | https://mcp.geonode.com/mcp |
| Transport | Streamable HTTP |
| Account authentication | X-Api-Key header |
| Official registry ID | com.geonode/mcp |
Connections to https://scraper.geonode.io/mcp keep working; it serves the same server.
extract runs without one; the other tools need your Geonode API key and use your plan limits.Client configuration formats differ. There is no universal JSON installation block in this repository, and the documented recipes are not a claim of tested installation in every client.
Requires Python 3.10 or newer; no third-party packages or API key are needed.
# Show what the example will do. No network calls.
python3 examples/discover.py
python3 examples/discover.py --connect
The script sends initialize, notifications/initialized, and tools/list. It handles JSON or SSE responses and preserves a session ID if the server provides one. It never sends tools/call, reads API credentials, or extracts a page. A successful result confirms metadata discovery, not scraping success.
How the discovery example works · Run its offline tests
Anonymous access is intended for limited synchronous extraction. JavaScript rendering and country targeting require authentication. Check the current keyless limits before relying on it for a workflow.
For authenticated access, create an API key in the Geonode dashboard and set X-Api-Key through your client's credential or secure-header configuration. Keep the key out of prompts and committed files. Use the authenticated tool schemas returned by your client rather than assuming the anonymous tool list is complete.
Expected result: The tool supplies page content and the assistant writes the answer. This repository supplies setup guides and a discovery example, not a model or a self-hosted scraping backend.
Use MCP when your assistant discovers and calls tools through the Model Context Protocol. Use the Python or TypeScript SDK when your application owns the request flow, or the CLI for shell scripts.
Use the native Make.com, Zapier or self-hosted n8n connector when building scenarios or workflows in those platforms. Their available operations and access requirements differ from the MCP tool list.
Anonymous access exposes limited extraction. For authenticated access, discover the actual tool schemas in your connected client instead of copying tool names or arguments from a framework toolkit.
These recipes describe source-supported workflows. They have not been run as authenticated end-to-end examples in this review.
| Resource | Purpose |
|---|---|
| Connection guide | Endpoint, authentication and troubleshooting |
| Discovery example | Inspect metadata and tool input schemas |
| Offline tests | Exercise JSON/SSE parsing and discovery flow |
| Source and release boundary | Separate this repository from the hosted backend |
The Python SDK, CLI and framework toolkits are separate integrations. Installing a Python package does not start this hosted MCP service. No local bridge, self-hosted backend, or package release is provided here.
For API or account questions, contact Geonode. For vulnerabilities, follow SECURITY.md. Contributions use the MIT license.
[Try Geonode for free](https://app.geonode.com/sign-up)