winpc-mcp
Enables remote control of a Windows PC from an iPad via MCP, providing tools for executing commands, managing files, taking screenshots, controlling input, and managing processes.
README
winpc-mcp
从 iPad(Minis)通过 MCP 远程控制 Windows 电脑。
在电脑上运行一个 MCP 服务器,iPad 上的 Minis 通过 HTTP 调用它的工具: 执行命令、管理文件、截屏、控制键盘鼠标、打开应用、管理进程、下载文件等 —— 就像在电脑本地跑了一个 agent。
┌────────── iPad ──────────┐ ┌────────── Windows 电脑 ──────────┐
│ Minis (MCP 客户端) │ HTTP │ winpc-mcp (MCP 服务器) │
│ minis-mcp-cli 调用工具 │◀──────▶│ :8765/mcp (Bearer Token) │
└──────────────────────────┘ └──────────────────────────────────┘
局域网直连 / Tailscale / 隧道
✨ 功能
| 类别 | 工具 |
|---|---|
| 命令 | run_powershell run_cmd run_shell |
| 文件 | read_file write_file list_dir search_files delete_file get_file_info |
| 屏幕 | screenshot get_screen_size |
| 输入 | type_text press_key click move_mouse scroll get_mouse_position get_clipboard set_clipboard |
| 应用 | open_app open_url list_windows focus_window close_window |
| 系统 | system_info list_process kill_process |
| 网络 | http_request download_file serve_file(电脑文件回传给 iPad) |
🚀 快速开始
1. 电脑端(Windows)
:: 方式一:一键脚本(自动建虚拟环境、装依赖、启动)
start.bat
:: 方式二:手动
python -m venv .venv
.venv\Scripts\activate
pip install -r requirements.txt
python server.py
启动后终端会打印:
MCP 地址: http://192.168.x.x:8765/mcp
Token: xxxxx(未配置时自动生成,请保存)
Dashboard: http://192.168.x.x:8765/dashboard
固定 Token:把
config.example.json复制为config.json,填入你的 token, 或设置环境变量WINPC_TOKEN。自动生成的 token 每次重启都会变。
2. iPad 端(Minis)
minis-mcp-cli add --name winpc \
--url http://<电脑IP>:8765/mcp \
--header "Authorization: Bearer <token>"
minis-mcp-cli tools winpc # 查看工具
minis-mcp-cli call winpc system_info # 测试调用
之后在 Minis 对话里直接说“帮我看看电脑的 CPU 占用”即可。
📊 Dashboard(电脑端实时监控)
电脑浏览器打开 http://<电脑IP>:8765/dashboard(或本机 localhost:8765):
- DIALOGUE:实时显示 iPad 端对话(Minis 通过
sync_chat工具同步,或POST /api/chat) - TOOL CALLS:每次工具调用的完整记录——参数、结果、耗时、成功/失败状态
- SCREEN + SYSTEM:最近截图画廊(点击放大)+ CPU/内存/磁盘实时状态
- ⌘K 命令面板:过滤工具调用、跳转面板、暂停/恢复实时流、清空日志、复制地址
可配置 dashboard_password 口令保护(开启后首次访问需输入口令)。
🔒 安全
- 所有
/mcp和/files请求必须带Authorization: Bearer <token> - 默认只监听局域网;不要把 8765 端口直接映射到公网
- 异地访问推荐 Tailscale(加密隧道),两端装 app 登录同一账号后,
MCP 地址直接换成 Tailscale IP(
http://100.x.x.x:8765/mcp)即可 config.json可配置:allowed_dirs:限定文件操作目录(空 = 不限制,完全操控)confirm_dangerous:true时删除文件 / 杀进程需显式confirm=true
⚙️ 配置
{
"host": "0.0.0.0", // 监听地址
"port": 8765, // 端口
"token": "", // MCP 认证 token(留空则自动生成)
"allowed_dirs": [], // 文件操作白名单目录(空=不限制)
"confirm_dangerous": false, // true=危险操作需确认
"dashboard_password": "" // Dashboard 访问口令(空=不设防)
}
📝 开发
# 本地(非 Windows)也能启动核心功能测试:
# run_powershell/run_cmd/键盘鼠标/截图等 Windows 专属工具会返回明确错误
pip install -r requirements.txt
python server.py
工具代码在 winpc/ 目录,新增工具 = 在 tools_*.py 里加一个带 @tool 装饰器和
类型注解 + docstring 的函数即可,启动时自动注册。
🛣️ 路线图
- [x] Dashboard:电脑浏览器实时查看对话、工具调用、截图
- [x] Dashboard 动效:状态过渡、计数反馈、清空退出(v1.1.1)
- [ ] 浏览器自动化(CDP 驱动 Chrome)
- [ ] 开机自启(任务计划程序)
📄 License
MIT
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.
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.
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.
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
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.
E2B
Using MCP to run code via e2b.