MCP Servers · mcpjungle

MCPJungle

MCPJungle is a self-hosted gateway that consolidates multiple Model Context Protocol (MCP) servers behind a single HTTP endpoint, eliminating the need to configure each server individually in AI clients like Claude or Cursor. It provides centralized tool discovery, access control, and observability for both personal and team deployments.

Talk to DEV.co Discuss Implementation

Source: GitHub — github.com/mcpjungle/MCPJungle

1.1k

GitHub stars

144

Forks

Primary language

MPL-2.0

License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

Field	Value
Repository	mcpjungle/MCPJungle
Owner	mcpjungle
Primary language	Go
License	MPL-2.0 — OSI-approved
Stars	1.1k
Forks	144
Open issues	92
Latest release	0.4.5 (2026-05-20)
Last updated	2026-05-20
Source	https://github.com/mcpjungle/MCPJungle

What MCPJungle is

Written in Go, MCPJungle exposes a streamable HTTP MCP server at `/mcp` that aggregates registered MCP servers (both HTTP and stdio-based), handles tool/prompt/resource discovery, and supports optional access control and OpenTelemetry instrumentation. It stores configuration persistently and can run in development or enterprise modes.

Quickstart

Get the MCPJungle source

Clone the repository and explore it locally.

terminalbash

1git clone https://github.com/mcpjungle/MCPJungle.git2cd MCPJungle3# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

Unified MCP endpoint for multi-client AI workflows

Replace scattered per-client MCP configurations with one gateway endpoint that Claude Desktop, Cursor, Copilot, and custom agents can connect to, reducing setup duplication.

Team-shared MCP infrastructure with centralized control

Deploy MCPJungle on shared infrastructure to give teams a single discoverable registry of tools, with centralized access control and observability instead of per-user setups.

Local development environment organization

Keep personal MCP server configurations clean and reproducible by registering all local and remote servers once in MCPJungle rather than managing them across multiple client config files.

Implementation considerations

Persistent PostgreSQL database is required; use provided docker-compose files for local/prod setup, or manage your own database instance.
Choose appropriate Docker image variant (standard vs. stdio-tagged) based on whether you register stdio-based servers that need `npx` or `uvx` runtime.
Use the CLI (`mcpjungle register`) to add servers by URL; stdio-based servers require additional configuration and the stdio image.
Access control and OpenTelemetry features are available in enterprise mode; enable via docker-compose.prod.yaml for team deployments.
Tool groups feature allows selective tool exposure to specific clients; requires configuration in MCPJungle but simplifies client-side security policies.

When to avoid it — and what to weigh

Single-client, single-server deployments — If you only use one AI client with one MCP server, direct client-to-server connection is simpler and avoids the added operational layer.
Stateless, ephemeral service requirements — MCPJungle requires persistent storage (PostgreSQL) and maintains connection state; if you need a purely stateless proxy, this adds unnecessary complexity.
Fully managed SaaS-only environments — MCPJungle is self-hosted only; if your organization requires managed hosting and SLA guarantees, you must operate and maintain the infrastructure yourself.
Extreme low-latency requirements — Adding a gateway layer introduces additional network hops; if sub-millisecond latency is critical, direct client-server connections are preferable.

License & commercial use

Licensed under Mozilla Public License 2.0 (MPL-2.0), a copyleft license requiring derivative works to disclose source code under the same license. Commercial use is permitted, but modifications must be released under MPL-2.0.

MPL-2.0 permits commercial use and proprietary services built on MCPJungle, provided modifications to MCPJungle itself are released under MPL-2.0. Running MCPJungle as an internal tool or offering it as a hosted service is allowed. Requires review if you modify MCPJungle core and wish to keep changes proprietary.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

Signal	Assessment
Maintenance	Active
Documentation	Strong
License clarity	Clear
Deployment complexity	Moderate
DEV.co fit	Good
Assessment confidence	High

Security considerations

No public security audit or vulnerability disclosure policy stated. Access control features available in enterprise mode but details not provided in README. Runs HTTP MCP server at localhost:8080 by default; network exposure must be managed deliberately. Tool groups provide granular tool visibility control. Recommend reviewing authentication mechanisms, TLS/encryption for remote deployments, and credential handling for registered server URLs. Consider security of persistent database containing server configurations.

Alternatives to consider

Direct client-to-server MCP connections

Simplest for single client/server pairs; no gateway overhead but requires per-client configuration and lacks centralized discovery/access control.

Anthropic's MCP Server (reference implementation)

Teaches MCP protocol but does not provide gateway functionality; appropriate for learning or building custom solutions, not for managing multiple servers.

Custom nginx/HAProxy reverse proxy

Lower-level alternative offering load balancing and TLS termination but lacks MCP protocol awareness, tool discovery, and access control features.

Software development agency

Build on MCPJungle with DEV.co software developers

Start with a local docker-compose setup today, then scale to team infrastructure with centralized access control and observability. Review the full docs at docs.mcpjungle.com.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

mcp mcp-server-browserbase mcp-server-kubernetes

Related on DEV.co

Explore the category and the services that help you build with it.

MCP Servers AI development API development DevOps & infrastructure Talk to DEV.co

MCPJungle FAQ

Can I use MCPJungle with Claude Desktop, Cursor, and other AI clients?

Yes. MCPJungle exposes a single HTTP MCP endpoint at `/mcp` that any client supporting HTTP MCP SSE transport can connect to. Claude Desktop, Cursor, and Copilot are explicitly supported via JSON config; custom agents can use the same HTTP endpoint.

Do I need a database to run MCPJungle?

Yes, MCPJungle requires a persistent PostgreSQL database. Docker Compose files include a containerized Postgres instance for convenience. You can also use a managed PostgreSQL service or self-hosted instance.

Can I register stdio-based MCP servers (e.g., filesystem, github)?

Yes, but requires the stdio-tagged Docker image (e.g., `latest-stdio`), which includes `npx` and `uvx` runtimes. Standard image is minimal and supports only HTTP-based servers.

What are tool groups and when should I use them?

Tool groups allow you to expose only specific tools to specific clients instead of all tools. Useful for team deployments where different roles should see different tools, or for limiting scope for security/compliance reasons.

Custom software development services

DEV.co helps companies turn open-source tools like MCPJungle into production software. Our software development services cover the full lifecycle — architecture, web development, integration, and maintenance — delivered by software developers and web developers who ship. Engage our software development agency to implement or customize it for your mcp servers stack.

Simplify your MCP infrastructure with MCPJungle

Start with a local docker-compose setup today, then scale to team infrastructure with centralized access control and observability. Review the full docs at docs.mcpjungle.com.

Talk to DEV.co Schedule a Consultation