SmoothOperatorAgentTools
Smooth Operator Agent Tools 支援強大的 Windows 自動化、瀏覽器控制以及 AI 驅動的電腦互動。無論您是要建構能夠操作電腦的 AI 代理、自動化業務流程,還是建立測試腳本,這些工具都能透過多種整合選項提供全面的解決方案。
Smooth Operator Agent Tools MCP Server
Smooth Operator Agent Tools provide robust capabilities for Windows automation, browser control, and AI-driven computer interaction. They are designed to empower AI agents to understand and utilize computer functionalities.
What it does:
This MCP server allows AI agents (like Claude) to interact with your Windows environment, control web browsers (specifically Chrome), perform UI automation, and execute keyboard/mouse actions. It also includes AI Vision capabilities for mouse control based on UI element descriptions.
How to use:
- Download the Windows App: Obtain the
Smooth Operator Agent ToolsWindows application. - Installation: Follow the setup instructions provided by the app.
- MCP Client Configuration: Configure your MCP client with the following
mcpServersentry to connect via stdio:{ "mcpServers": { "SmoothOperatorAgentTools": { "command": "C:\\Users\\[USERNAME]\\AppData\\Roaming\\SmoothOperator\\AgentToolsServer\\smooth-operator-server.exe", "args": [ "/silent", "/close-with-parent-process" ] } } } - HTTP Server (Optional): The tools also expose functionality via an HTTP server at
http://localhost:54321, which can be used as a streamable endpoint. - Authentication: All API requests require an API key as a Bearer token in the
Authorizationheader. This key is shared with Screengrasp and can be obtained fromhttps://screengrasp.com/api.html.
Key Features:
- Screenshot and Analysis: Capture screenshots and analyze system state, UI elements, and application details.
- Mouse Control: Precise coordinate-based and AI Vision-powered mouse operations.
- Keyboard Input: Typing text, hotkeys, and key combinations.
- Chrome Browser Control: Navigation, DOM manipulation, and JavaScript execution.
- Windows Automation: Advanced UI Automation and code execution.
- API Documentation: Endpoints for accessing API documentation directly.
Implementation Details:
The server is provided as a Windows executable (.exe). Client libraries are available for Python, TypeScript, and C#/.NET, indicating a multi-language support for integration.
Recommend MCP Servers 💡
MCPControl
Windows control server for the Model Context Protocol, providing programmatic control over system operations including mouse, keyboard, window management, and screen capture functionality
fetcher-mcp
MCP server for fetching web page content using Playwright headless browser.
@mobilenext/mobile-mcp
Model Context Protocol Server for Mobile Automation and Scraping (iOS, Android, Emulators, Simulators and Real Devices)
playwright
Playwright MCP server
facebook-mcp-server
Facebook MCP server for automating posts, comment moderation, insights, and sentiment filtering.
Tabby MCP
A Tabby plugin that implements a Model Context Protocol (MCP) server, enabling AI-powered control and automation of the Tabby terminal.