MCPJungle: Self-Hosted MCP Gateway for AI Agents and Tool Management

Summary

MCPJungle is an open-source, self-hosted Model Context Protocol (MCP) Gateway designed for managing AI agents and their tool-calling capabilities. It allows developers and organizations to centralize the registration, discovery, and consumption of MCP servers and their tools, enhancing security and control over AI agent interactions.

Use at your own risk

OSRepos shares public repositories for knowledge and discovery only. Any installation, execution, configuration, or use of code from these repositories is the user's own responsibility. Always review the repository, source code, dependencies, licenses, and security implications before running or installing anything. OSRepos is not responsible for issues, damages, or losses resulting from third-party repositories.

Introduction

MCPJungle is an open-source, self-hosted Model Context Protocol (MCP) Gateway built in Go, designed to centralize the management of AI agents and their tool-calling capabilities. It acts as a unified gateway for your private AI agents, allowing you to register and manage various MCP servers and the tools they provide from a single, secure location. This enhances security, privacy, and access control for your AI agent interactions.

Installation

Getting started with MCPJungle is straightforward. You can quickly set up the server using Docker Compose or install the CLI tool directly.

To start the MCPJungle server locally with Docker Compose:

curl -O https://raw.githubusercontent.com/mcpjungle/MCPJungle/refs/heads/main/docker-compose.yaml
docker compose up -d

For managing MCPJungle from your local machine, install the CLI using Homebrew:

brew install mcpjungle/mcpjungle/mcpjungle

You can also pull the Docker image for the server:

docker pull ghcr.io/mcpjungle/mcpjungle

Examples

MCPJungle simplifies the process of integrating and managing tools for your AI agents. Here are a few examples:

Registering an MCP Server: You can register a streamable HTTP-based MCP server, like context7, using the CLI:

mcpjungle register --name context7 --url https://mcp.context7.com/mcp

Connecting AI Clients: Configure your AI clients, such as Claude or Cursor, to connect to MCPJungle. For Claude, you would add the following to your mcpServers configuration:

{
  "mcpServers": {
    "mcpjungle": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "http://localhost:8080/mcp",
        "--allow-http"
      ]
    }
  }
}

For Cursor, the configuration is even simpler:

{
  "mcpServers": {
    "mcpjungle": {
      "url": "http://localhost:8080/mcp"
    }
  }
}

Once connected, your AI agents can discover and utilize the tools registered in MCPJungle.

Tool Groups for Granular Control: MCPJungle allows you to create Tool Groups, exposing only a specific subset of tools to your AI clients. This helps manage performance and provides granular access control. For example, to create a group with specific tools:

{
  "name": "claude-tools",
  "description": "This group only contains tools for Claude Desktop to use",
  "included_tools": [
    "filesystem__read_file",
    "deepwiki__read_wiki_contents",
    "time__get_current_time"
  ]
}

Then, create the group using the CLI:

mcpjungle create group -c ./claude-tools-group.json

Why Use MCPJungle

MCPJungle addresses critical needs for developers and organizations working with AI agents:

For Developers using MCP Clients: It provides a central hub to access MCP servers for tool-calling, simplifying integration with clients like Claude and Cursor.
For Developers Building Production-Grade AI Agents: It offers built-in security, privacy, and Access Control, essential for robust and reliable AI agent deployments.
For Organizations: It enables centralized viewing and management of all MCP client-server interactions, hosted securely within their own datacenter, ensuring data governance and operational oversight.