pathintegral-institute/mcpm.sh

What it solves

MCPM simplifies the management of Model Context Protocol (MCP) servers. Instead of configuring servers individually for every AI client, it provides a global configuration model where servers are installed once and can be organized into profiles and then integrated into various MCP-compatible clients like Claude Desktop, Cursor, and Windsurf.

How it works

MCPM acts as a central manager for MCP servers. It allows users to search for servers via a central registry, install them globally, and organize them using "virtual profiles" (tags) to group servers for specific workflows. It can execute servers over stdio or HTTP for testing and can automatically update servers based on their source (Git, NPX, or binary).

Who it’s for

Developers and AI power users who use multiple MCP-compatible AI clients and want a single place to manage, discover, and share their server configurations without repetitive manual setup.

Highlights

• Global Server Management: Install servers once and deploy them across multiple clients.
• Server Discovery: Integrated search and installation from the MCP Registry.
• ** uma Virtual Profiles**: Organize servers into logical groups for different tasks.
• Client Integration: Direct configuration management for clients like Claude Desktop, Cursor, Windsurf, and VS Code.
• AI Agent Friendly: Includes a non-interactive CLI mode and an llm.txt guide for automation by AI agents.
• Remote Sharing: Ability to share local servers through secure tunnels for remote access.