MCPHub — Centralized MCP Server Management Hub

MCPHub — Unified MCP Server Management

The Problem

As MCP adoption grows, teams end up managing dozens of MCP servers — each with its own configuration, endpoint, and auth. Keeping track of which server handles what becomes a management nightmare.

The Solution

MCPHub acts as a gateway between your AI agents and your MCP servers. One dashboard to manage them all, with intelligent routing to pick the right server for each request.

Key Features

• Centralized dashboard — monitor all MCP servers from one place
• Smart routing — AI-powered vector semantic search routes requests to the right server
• Flexible routing strategies — route to all servers, specific groups, or individual servers
• Hot-swap configuration — update server configs without downtime
• OAuth 2.0 authentication — secure access to your MCP endpoints
• PostgreSQL integration — database mode for persistent configuration
• Docker deployment — one command to get started

Setup

# Quick start
docker run -p 3000:3000 samanhappy/mcphub

# With custom config
docker run -p 3000:3000 
  -v ./mcp_settings.json:/app/mcp_settings.json 
  samanhappy/mcphub

Configuration Example

{
  "servers": [
    {
      "name": "github",
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-github"]
    },
    {
      "name": "postgres",
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-postgres"]
    }
  ],
  "routing": {
    "strategy": "smart",
    "fallback": "round-robin"
  }
}

Routing Strategies

FAQ

Q: What is MCPHub? A: MCPHub is a centralized management platform that organizes multiple MCP servers into unified endpoints with smart routing, monitoring, and hot-swap configuration.

Q: Is MCPHub free? A: Yes, fully open-source under Apache-2.0. Self-hosted via Docker.

Q: How does smart routing work? A: MCPHub uses vector semantic search to analyze incoming requests and route them to the MCP server best equipped to handle them, with fallback strategies for reliability.