Discover Awesome MCP Servers

GitHub MCP Server Plus

Mirror of

pty-mcp

An MCP tool server that provides a stateful terminal.

browser-use MCP Server

Un servidor MCP basado en TypeScript que implementa un sistema de notas sencillo, permitiendo a los usuarios crear, acceder y generar resúmenes de notas de texto a través de Claude Desktop.

Solana Model Context Protocol (MCP) Demo

Una implementación sencilla de un servidor MCP que proporciona a los modelos de IA métodos RPC básicos de Solana y avisos de desarrollo útiles.

LMStudio-MCP

Un puente que permite a Claude comunicarse con modelos LLM que se ejecutan localmente a través de LM Studio, permitiendo a los usuarios aprovechar sus modelos privados a través de la interfaz de Claude.

Swift MCP GUI Server

There isn't a single, widely known "MCP server" specifically designed for executing keyboard input and mouse movement commands on macOS. The term "MCP server" is quite general. However, you can achieve this functionality using a combination of technologies and approaches. Here are a few options, ranging from simpler to more complex: **1. Using AppleScript and a Simple Server (Relatively Simple):** * **AppleScript:** macOS has powerful scripting capabilities through AppleScript. You can write AppleScript code to simulate keyboard input and mouse movements. * **Simple Server (e.g., Python with Flask or a simple TCP server):** You can create a simple server (e.g., using Python with the Flask framework or a basic TCP socket server) that listens for commands. When it receives a command, it executes the corresponding AppleScript. * **Example (Conceptual):** 1. **Server (Python with Flask):** ```python from flask import Flask, request import subprocess app = Flask(__name__) @app.route('/execute_applescript', methods=['POST']) def execute_applescript(): script = request.form['script'] try: subprocess.run(['osascript', '-e', script], check=True) return "AppleScript executed successfully", 200 except subprocess.CalledProcessError as e: return f"Error executing AppleScript: {e}", 500 if __name__ == '__main__': app.run(debug=True, host='0.0.0.0') # Make sure to understand the security implications of this ``` 2. **AppleScript (Example - Move Mouse):** ```applescript tell application "System Events" set the position of the mouse to {100, 100} end tell ``` 3. **Client (Sending the Command):** You'd use a client (e.g., `curl`, `requests` in Python, etc.) to send a POST request to the Flask server with the AppleScript code in the `script` parameter. * **Security Considerations:** This approach has significant security implications. Anyone who can access the server can execute arbitrary AppleScript code on your macOS system. **Do not expose this server to the public internet without proper authentication and authorization.** Consider using a more robust authentication mechanism than just relying on the server being on a private network. **2. Using `osascript` Directly with SSH (Simple, but Limited):** * You can use SSH to remotely execute `osascript` commands. This avoids the need for a separate server process. * **Example:** ```bash ssh user@mac_address "osascript -e 'tell application \"System Events\" to key code 125'" # Down arrow key ``` * **Limitations:** This is less flexible than a dedicated server, as you need to construct the full `osascript` command on the client side. It's also less suitable for complex interactions. **3. Using a Remote Desktop Protocol (RDP) or Virtual Network Computing (VNC) Server (More Robust):** * **RDP (e.g., using FreeRDP or similar):** While macOS doesn't natively support RDP as a server, you can install third-party RDP server software. This allows you to remotely control the entire desktop, including keyboard and mouse. * **VNC (e.g., using macOS's built-in Screen Sharing or a third-party VNC server like RealVNC):** VNC is a widely used protocol for remote desktop access. macOS has a built-in VNC server (Screen Sharing). * **Advantages:** Full remote desktop control, including keyboard and mouse. * **Disadvantages:** Higher overhead than simpler solutions. Requires a VNC or RDP client on the controlling machine. **4. Using a Custom Daemon/Service with Accessibility API (More Complex, Most Control):** * **Create a Daemon/Service:** Develop a background process (daemon or service) that runs on macOS. * **Accessibility API:** Use the macOS Accessibility API to programmatically control the keyboard and mouse. This requires enabling accessibility permissions for your application in System Preferences. * **Communication:** Implement a communication protocol (e.g., TCP sockets, gRPC, etc.) for the daemon to receive commands from a remote client. * **Advantages:** Fine-grained control, can be highly optimized. * **Disadvantages:** Most complex to implement. Requires understanding of macOS system programming and the Accessibility API. Security is paramount; carefully consider how you authenticate and authorize commands. **Code Example (AppleScript - Keyboard Input):** ```applescript tell application "System Events" keystroke "Hello, world!" key code 125 -- Down arrow key key code 36 -- Return key end tell ``` **Code Example (AppleScript - Mouse Click):** ```applescript tell application "System Events" click at {100, 100} -- Click at coordinates (100, 100) end tell ``` **Important Considerations:** * **Security:** Remote control of keyboard and mouse is a sensitive operation. Implement robust authentication and authorization to prevent unauthorized access. Never expose such a service to the public internet without proper security measures. * **Accessibility Permissions:** Controlling keyboard and mouse programmatically often requires enabling accessibility permissions for your application in System Preferences > Security & Privacy > Accessibility. The user must explicitly grant these permissions. * **Error Handling:** Implement proper error handling in your server and client code to gracefully handle unexpected situations. * **Concurrency:** If you expect multiple clients to connect to your server, consider using threading or asynchronous programming to handle concurrent requests. * **macOS Updates:** macOS updates can sometimes break compatibility with applications that use the Accessibility API or other system-level features. Be prepared to update your code as needed. **Which approach is best?** * For simple, occasional tasks, using SSH with `osascript` might be sufficient. * For more complex control and a dedicated server, the AppleScript/Flask approach (with strong security) or a custom daemon with the Accessibility API are better choices. * If you need full remote desktop access, RDP or VNC are the most appropriate options. Remember to thoroughly research and understand the security implications of each approach before implementing it. Start with the simplest solution that meets your needs and add complexity only if necessary.

Mcp Allure Server

Un servidor que convierte los informes de prueba de Allure a formatos compatibles con modelos de lenguaje grandes (LLM), lo que permite a los modelos de IA analizar mejor los resultados de las pruebas y proporcionar información sobre los fallos de las pruebas y posibles soluciones.

PancakeSwap PoolSpy MCP Server

Un servidor MCP que rastrea los pools de liquidez recién creados en Pancake Swap.

Linear MCP Server

linear MCP server based on mcp-go

Patent_mcp_server

Servidor FastMCP para datos de la USPTO

MCP GitHub Repository Server

Mirror of

Snak

Cree agentes de IA potentes y seguros impulsados por Starknet.

Pica Mcp Server

A Model Context Protocol Server for Pica, built in TypeScript

Internetsearch-mcp-server

Un servidor MCP para búsqueda en internet, basado en la API de búsqueda de Bocha.

weather-service MCP server

Espejo de

mcp-server-oracle

Servidor de Protocolo de Contexto de Modelo para acceder a la base de datos Oracle.

UseGrant MCP Server

Un servidor de Protocolo de Contexto de Modelo que proporciona herramientas para la gestión de proveedores, clientes, inquilinos y tokens de acceso a través de la plataforma UseGrant.

MCP Server Whisper

Un servidor MCP para la transcripción de audio utilizando OpenAI.

Steel Puppeteer

Espejo de

MCP Server Template

Una plantilla basada en TypeScript para desarrollar servidores del Protocolo de Contexto de Modelo con características como inyección de dependencias y arquitectura basada en servicios, facilitando la creación e integración de herramientas personalizadas de procesamiento de datos.

MCP Server for Executing Terminal Commands

Un servidor MCP que permite a asistentes de IA como Claude ejecutar comandos de terminal en la computadora del usuario y devolver la salida, funcionando como una terminal a través de la IA.

MCP Server Boilerplate

MCP Server Boilerplate is a Node.js example that demonstrates how to build an MCP server with a custom tool, including one that loads an environment variable from a configuration file, to integrate seamlessly with AI-assisted environments like Cursor AI.