How do I use Windows-MCP?

1. Click on "Install Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Windows-MCP open Notepad and type 'hello world'" That's it! The server will respond to your query, and you can continue using it as needed. Here is a step-by-step guide with screenshots.

Windows-MCP

by Books-QAQ

Overview Schema Related Servers Score Discussions

Python

Local

Windows-MCP 是一个轻量级开源 MCP 服务器，桥接 AI Agent 与 Windows 操作系统，支持文件操作、应用控制、UI 交互、自动化测试等任务。

✨ 核心特性

无缝 Windows 集成 — 原生操控 Windows UI 元素、启动应用、控制窗口、模拟用户输入。
任意 LLM 驱动 — 不依赖计算机视觉或微调模型，兼容任何 LLM 提供商。
17 个 MCP 工具 — Click、Type、Scroll、Move、Shortcut、Screenshot、Snapshot、App、Shell、Scrape、MultiSelect、MultiEdit、Clipboard、Process、Notification、Registry、Wait。
A2A 编排框架 — LLM 任务规划、DAG 并行执行、结构化技能输出、多 Agent 协作委托。
轻量开源 — 最少依赖。

Related MCP server: AETHER_01

🧠 A2A 框架

Agent-to-Agent 编排层：

组件	说明
技能系统	5 个内置技能（打开应用、截图、搜索、文件操作、剪贴板），插件化架构 + JSON Schema 约束
DAG 执行器	拓扑排序、同层并行执行、跨节点上下文传递（`$step_id.field`）
LLM 规划器	自然语言 → `TaskGraph` JSON，支持 OpenAI 兼容 API，自动降级关键词匹配
智能验证器	两层校验：规则检查（进程/文件/剪贴板）+ 可选视觉模型对比
A2A 网络	`AgentCard` 能力发现、`AgentRegistry` 注册中心、FastAPI 远程任务委托

使用示例

from windows_mcp.mobile.runtime import create_mobile_runtime

agent, skills = create_mobile_runtime()

# 简单指令（关键词匹配）
agent.run_instruction("打开QQ")

# DAG 顺序链
agent.run_sequential([
    {"skill": "open_or_focus_app", "params": {"instruction": "打开Chrome"}},
    {"skill": "capture_desktop_state"},
])

# LLM 智能规划（需配置 PLANNER_API_KEY 环境变量）
agent.run_instruction_smart("打开浏览器搜索Python教程然后截图保存")

🛠️ 安装

首次安装依赖解析约需 1–2 分钟。

环境要求

Python 3.13+
UV 包管理器

方式 A：PyPI 安装（推荐）

{
  "mcpServers": {
    "windows-mcp": {
      "command": "uvx",
      "args": ["windows-mcp"]
    }
  }
}

方式 B：源码安装

git clone https://github.com/Books-QAQ/Windows-MCP.git
cd Windows-MCP

{
  "mcpServers": {
    "windows-mcp": {
      "command": "uv",
      "args": ["--directory", "<项目路径>", "run", "windows-mcp"]
    }
  }
}

MSIX 版本（Windows Store）：配置文件位于 %LOCALAPPDATA%\Packages\Claude_pzs8sxrjxfjjc\LocalCache\Roaming\Claude\。如 Electron 未继承 PATH，请使用 uv.exe 的完整绝对路径。

# PyPI
claude mcp add --transport stdio windows-mcp -- uvx windows-mcp

# 源码
git clone https://github.com/Books-QAQ/Windows-MCP.git
claude mcp add --transport stdio windows-mcp -- uv --directory "<路径>" run windows-mcp

{
  "mcpServers": {
    "windows-mcp": {
      "command": "uvx",
      "args": ["windows-mcp"]
    }
  }
}

源码安装：将 "command": "uvx" 替换为 "command": "uv"，"args": ["--directory", "<路径>", "run", "windows-mcp"]。

🖥️ 运行模式

本地模式（默认）

uvx windows-mcp                           # stdio（默认）
uvx windows-mcp --transport sse --host localhost --port 8000
uvx windows-mcp --transport streamable-http --host localhost --port 8000

远程模式

需要在 windowsmcp.io 注册账号。

环境变量配置：

MODE=remote
SANDBOX_ID=<沙箱ID>
API_KEY=<API密钥>

⚙️ 环境变量

A2A 规划器与验证器

变量	默认值	说明
`PLANNER_API_KEY`	无	LLM 任务规划的 API Key
`PLANNER_BASE_URL`	`https://api.deepseek.com/v1`	LLM API 地址
`PLANNER_MODEL`	`deepseek-chat`	规划模型名称
`VALIDATOR_API_KEY`	无	视觉验证的 API Key
`VALIDATOR_MODEL`	`deepseek-chat`	视觉验证模型

截图与快照

变量	默认值	说明
`WINDOWS_MCP_SCREENSHOT_SCALE`	`1.0`	缩放比例（0.1–1.0），4K 屏建议 `0.5` 以避免超过 1MB 限制
`WINDOWS_MCP_SCREENSHOT_BACKEND`	`auto`	截图后端：`auto` / `dxcam` / `mss` / `pillow`
`WINDOWS_MCP_PROFILE_SNAPSHOT`	关闭	设为 `1`/`true` 启用性能计时日志

遥测与调试

变量	默认值	说明
`ANONYMIZED_TELEMETRY`	`true`	设为 `false` 关闭匿名遥测
`WINDOWS_MCP_DEBUG`	`false`	设为 `1`/`true` 启用调试日志

🔨 MCP 工具列表

工具	功能
Click	按坐标或 UI 元素标签点击
Type	输入文本，支持清空/回车
Scroll	垂直/水平滚动
Move	移动鼠标或拖拽
Shortcut	键盘快捷键（`Ctrl+C`、`Alt+Tab` 等）
Wait	等待指定时长
Screenshot	快速截图，含光标和窗口信息
Snapshot	完整桌面状态，含交互元素和 DOM 提取
App	启动/缩放/切换应用
Shell	执行 PowerShell 命令
Scrape	抓取网页内容
MultiSelect	多选文件/复选框
MultiEdit	多字段同时输入
Clipboard	读写剪贴板
Process	列出或终止进程
Notification	发送 Windows 通知
Registry	读写注册表

📝 已知限制

无法选中段落中的特定文本（依赖无障碍树）。
Type 工具适用于文本输入，不适合 IDE 编程场景。
不适用于游戏自动化。

This server cannot be installed

RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

Maintenance

ActivityMaintained

ResponsivenessSyncing

Resources

GitHub Repository

Need Help?

Related Servers

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

Frenchie A
io.github.Lab94
OCR, transcription, file extraction, and image generation for AI agents via MCP.
AI & Machine Learning Image & Video Processing Speech Processing
72
Airbyte Agents
io.github.airbytehq
Let AI agents query data and act across all your business apps via MCP.
Data Platforms AI & Machine Learning RAG Systems
2
agentdocs
io.github.dancourse
An agent-first office suite Claude & ChatGPT read and write over one MCP URL.
Workplace & Productivity Documentation Access
2
anotherai
dev.anotherai
MCP server for building and testing AI agents with multi-model experimentation and insights.
Agent Orchestration Autonomous Agents Testing & QA Tools
2

Related MCP Servers

OmniFlow
OS Automation App Automation
huppugo1
A
license
Not graded
quality
D
maintenance
Enables AI clients to automate Windows desktop applications through window manipulation, image recognition, OCR, keyboard/mouse simulation, and memory operations via the MCP protocol.
Updated 3 months ago (2026-06-13 09:57 UTC)
MIT
AETHER_01
OS Automation Shell Access File Systems
Githubredminote105g
A
license
Not graded
quality
C
maintenance
Enables AI agents to perform extensive Windows system administration, file operations, process management, network configuration, registry editing, GUI automation, and more through a comprehensive set of MCP tools.
Updated 8 months ago (2026-01-02 13:01 UTC)
1
MIT
Windows-MCP
OS Automation Browser Automation File Systems
Vinayreddy18
A
license
Not graded
quality
C
maintenance
A lightweight MCP server that enables AI agents to seamlessly integrate with Windows, performing tasks like file navigation, application control, UI interaction, and QA testing.
Updated 14 days ago (2026-08-20 18:17 UTC)
1
MIT
Windows-MCP
OS Automation File Systems Shell Access
islamulhaq2811
A
license
Not graded
quality
B
maintenance
Enables AI agents to interact with the Windows operating system, performing tasks such as file navigation, application control, UI interaction, and QA testing.
Updated 14 days ago (2026-08-21 02:35 UTC)
MIT

Latest Blog Posts

Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)
By Om-Shree-0709 on July 25, 2026.
mcp
Agent Identity
OAuth 2.1
Your AI Chatbot Just Exposed Your CEO's Salary to an Intern
By Om-Shree-0709 on July 2, 2026.
Agent Identity
MCP Security
OAuth Delegation
Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)
By Om-Shree-0709 on June 30, 2026.
Agentic Ai
Prompt Injection
WebAssembly

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/Books-QAQ/Windows-MCP'

If you have feedback or need assistance with the MCP directory API, please join our Discord server

✨ 核心特性

🧠 A2A 框架

使用示例

🛠️ 安装

环境要求

🖥️ 运行模式

本地模式（默认）

远程模式

⚙️ 环境变量

A2A 规划器与验证器

截图与快照

遥测与调试

🔨 MCP 工具列表

📝 已知限制

Maintenance

Resources

Looking for Admin?

Related MCP Connectors

Related MCP Servers

OmniFlow

AETHER_01

Windows-MCP

Windows-MCP

Latest Blog Posts

MCP directory API