Pluggedin

The plugged.in MCP Proxy Server is a middleware that aggregates multiple MCP servers to provide a unified interface, supports multiple MCP clients, and has advanced management functions.

Developer tools Artificial intelligence chatbots #AI middleware #Protocol proxy #Unified interface #Multi - server Local .TypeScript

rating : 2.5 points

downloads : 11

update time : 2025-04-28

What is the plugged.in MCP Proxy Server?

The plugged.in MCP Proxy Server is an intelligent middleware that can aggregate multiple Machine Conversation Protocol (MCP) servers into a single interface. By integrating with the plugged.in application, it automatically retrieves tool, prompt, and resource configurations and intelligently routes requests to the appropriate underlying MCP servers.

How to use the plugged.in MCP Proxy Server?

Simply install the Node.js environment and obtain a plugged.in API key, then you can start the proxy server with a simple command. It supports seamless integration with MCP clients such as Claude Desktop, Cline, and Cursor.

Use cases

It is suitable for scenarios where multiple MCP servers need to be managed simultaneously, especially in scenarios where a unified interface and data exchange are required between different AI tools. Typical use cases include AI development environments, automated workflows, and multi - model collaboration systems.

Main features

Universal MCP compatibilitySupports all major MCP clients, including Claude Desktop, Cline, and Cursor

Multi-server supportCan connect to both STDIO (command - line) and WebSocket (HTTP) types of MCP servers simultaneously

Namespace isolationKeeps tools and resources of different MCP servers isolated and organized through prefixes

Multi-workspace layerSwitch between different MCP configuration sets with one click to meet the needs of different projects

API-driven proxyObtains capability configurations from the plugged.in application API instead of direct discovery

Advantages and limitations

Advantages

Simplifies the management of multiple MCP servers and provides a unified interface

Supports seamless integration of major MCP clients

Centrally manages configurations through the API, reducing maintenance complexity

Docker support facilitates deployment and expansion

An open - source project driven by the community

Limitations

Requires a Node.js runtime environment

Depends on the plugged.in API service

Requires obtaining an API key for initial configuration

Has limited support for custom MCP extensions

How to use

Installation preparation

Ensure that the Node.js 18+ environment is installed. It is recommended to use Node 20+

Obtain an API key

Visit the plugged.in website to register and obtain an API key

Start the proxy server

Use npx to quickly start the proxy server

Configure the MCP client

Add proxy configurations according to the client (Claude/Cline/Cursor) you are using

Usage examples

Example 1: Integrate multiple AI toolsUse the AI capabilities of Claude and Cursor simultaneously and call them through a unified interface

Example 2: Switch between multiple workspacesConfigure different sets of MCP servers for different projects

Example 3: Centralized resource managementAccess resources distributed on different servers through the proxy

Frequently Asked Questions

Do I need to configure each MCP server separately?

Which MCP clients are supported?

How to update the server configuration?

Is local development and debugging supported?

Related resources

plugged.in Application Repository

Source code and documentation of the plugged.in application

MCP Protocol Specification

Official documentation of the Machine Conversation Protocol

Claude Documentation

User guide for Claude Desktop

Docker Installation Guide

Official Docker installation documentation

🚀 plugged.in MCP Proxy Server

The plugged.in MCP Proxy Server is a tool designed to bridge multiple Machine Conversation Protocol (MCP) servers, enabling MCP clients to discover and utilize features, resources, templates, and prompts across different MCP servers.

🚀 Quick Start

✨ Features

Establishes a bridge between various MCP servers.
Enables MCP clients to discover and use distributed features, resources, templates, and prompts.
Integrates seamlessly with the pluggedin-app.
Provides a web - based interface for MCP server configuration management.
Supports centralized discovery of functions, custom instruction management, multi - workspace configuration, an interactive console for testing MCP tools, and user authentication and API key management.

📦 Installation

To install the plugged.in MCP Proxy Server, follow these steps:

git clone https://github.com/yourusername/pluggedin-mcp-proxy.git
cd pluggedin-mcp-proxy
npm install

Configuration

Create a config.json file and add the following content:

{
  "port": 3001,
  "appUrl": "http://localhost:3000",
  "servers": [
    {
      "name": "local-server",
      "host": "localhost",
      "port": 3002,
      "workspace": "default"
    },
    {
      "name": "remote-server",
      "host": "api.example.com",
      "port": 8080,
      "workspace": "production"
    }
  ]
}

🔧 Technical Details

Workflow

Configuration: The proxy server retrieves server configurations from the plugged.in application.
Function List: The proxy server fetches discovered functions via APIs:
- tools/list: Obtained from /api/tools (returns prefixed names).
- resources/list: Obtained from /api/resources.
- resource-templates/list: Obtained from /api/resource-templates.
- prompts/list: Merged results from /api/prompts and /api/custom-instructions.
Function Resolution: The proxy server resolves functions to target servers:
- tools/call: Parses the prefix in the tool name and looks up the server in the internal mapping.
- resources/read: Calls /api/resolve/resource?uri=... to get server details.
- prompts/get: Checks for custom instruction prefixes or calls /api/resolve/prompt?name=....
Request Routing: Requests are routed to the appropriate underlying MCP servers.
Response Handling: Responses from the underlying servers are passed to the clients.