AutoFlow
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@AutoFlowfind and click the 'Sign In' button on the login screen"
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.
AutoFlow
Windows 自动化工具,快速编辑工作流,完成日常重复性任务。 —— 截图识别 + 键鼠模拟,用可视化流程编辑器拖拽编排自动化任务。
附带实现 MCP Server,也可以让 AI Agent 直接操控桌面。
它能做什么
看屏幕:在屏幕上找到需要的位置(OpenCV 模板匹配)
控键鼠:点击、移动、打字、按键、滚动
编排流程:拖拽式流程编辑器,支持条件分支、循环、变量
MCP 接入:18 个 MCP tools,AI Agent 也能用
Related MCP server: Windows MCP Server
快速开始
# 后端
uv sync
uv run uvicorn app.main:app --reload
# 前端
cd webui && npm install && npm run dev后端 http://127.0.0.1:8000,前端 http://localhost:5173(自动代理 API)。
架构
三层分离,engine 层纯 Python 零框架依赖:
Engine (app/engine/) ← 纯业务逻辑,可独立测试
↑
Core API (app/api/ + app/ws/ + app/mcp/) ← REST / WebSocket / MCP 三种接入
↑
WebUI (webui/) ← React + ReactFlow 可视化编辑器引擎层用抽象基类,实现可替换:
截图:
ScreenCapture→ MSS找图:
ImageMatcher→ OpenCV 模板匹配输入:
InputController→ PyAutoGUI
MCP Server
如果你想让 AI Agent 接入,有两种方式:
HTTP — 启动后端后自动可用,端点 http://127.0.0.1:8000/mcp
stdio — 配置到 Claude Desktop / Claude Code:
{
"mcpServers": {
"autoflow": {
"command": "uv",
"args": ["run", "python", "-m", "app.mcp"],
"cwd": "/path/to/AutoFlow"
}
}
}技术栈
后端 Python 3.12+ / FastAPI / SQLModel + SQLite / MCP SDK
引擎 mss / OpenCV / pyautogui
前端 React 19 / TypeScript / ReactFlow / Vite
工具链 uv / Ruff / pytest
Roadmap
引擎层抽象 + 实现
REST API + WebSocket 实时日志
可视化流程编辑器(拖拽、条件、循环、变量)
MCP Server(HTTP + stdio)
内置截图选区工具
桌面打包(pywebview + PyInstaller)
驱动级键鼠(Interception)
高级图像识别(DINOv2)
License
MIT
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Latest Blog Posts
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
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/66maer/AutoFlow'
If you have feedback or need assistance with the MCP directory API, please join our Discord server