MCP Servers · samanhappy

mcphub

MCPHub is a TypeScript-based orchestration platform that centralizes management of multiple Model Context Protocol (MCP) servers through a unified HTTP/SSE gateway. It provides a web dashboard, CLI tools, flexible routing, and PostgreSQL support for production deployments.

Talk to DEV.co Discuss Implementation

Source: GitHub — github.com/samanhappy/mcphub

2.2k

GitHub stars

270

Forks

TypeScript

Primary language

Apache-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	samanhappy/mcphub
Owner	samanhappy
Primary language	TypeScript
License	Apache-2.0 — OSI-approved
Stars	2.2k
Forks	270
Open issues	32
Latest release	v1.0.23 (2026-07-05)
Last updated	2026-07-08
Source	https://github.com/samanhappy/mcphub

What mcphub is

Node.js/Express backend with React frontend that routes MCP server connections via HTTP/SSE endpoints, supports OAuth 2.0 and bearer authentication, uses file-based or PostgreSQL configuration with pgvector for semantic search, and includes hot-swappable server management without downtime.

Quickstart

Get the mcphub source

Clone the repository and explore it locally.

terminalbash

1git clone https://github.com/samanhappy/mcphub.git2cd mcphub3# follow the project's README for install & configuration

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

Best use cases

Multi-tool AI orchestration

Aggregate multiple MCP servers (fetch, time, compute, etc.) into a single gateway for AI clients (Claude, Cursor) to access via unified endpoints with smart routing and tool discovery.

Production MCP infrastructure

Operate MCP servers at scale with PostgreSQL backing, OAuth 2.0 authentication, social login, hot-reload configuration, and granular per-server visibility controls in a containerized environment.

Internal tool marketplace

Host an internal discovery API and public marketplace for MCP servers, allowing teams to share, install, and manage tools via CLI or dashboard without direct server-to-server wiring.

Implementation considerations

PostgreSQL is optional for file-based mode but required for production database backing, social login, and smart routing features; plan database provisioning and pgvector extension setup.
Authentication defaults to enabled on MCP endpoints; configure Bearer tokens or OAuth flows before exposing to external clients; admin password should be changed from generated default on first login.
Docker image variants (standard vs. full with Rust/Playwright) must match target MCP servers' runtime dependencies; mismatched variant may cause server launch failures.
Hot-swappable configuration requires the hub service to remain running; monitor service availability as all client connections depend on it; no built-in HA/failover mentioned.
CLI management requires network access back to hub API endpoint; CI/CD integration needs pre-configured authentication keys and endpoint URL passed to `mcphub` binary.

When to avoid it — and what to weigh

Minimal operational overhead required — MCPHub adds a persistent service layer. If you need zero additional infrastructure or direct peer-to-peer MCP routing, direct client-to-server connections may be simpler.
Single-server, ad-hoc MCP use — For one-off testing of a single MCP server or prototyping, the overhead of deployment, configuration, and authentication may outweigh benefits. Direct CLI invocation is faster.
Closed-source proprietary deployment required — MCPHub is Apache 2.0. If your workflow requires patent protection or closed-source modifications without source disclosure, licensing constraints apply; requires review by legal.
Air-gapped environments without external container registry access — MCPHub relies on Docker Hub for image distribution and runtime access to npm/pip/uvx package managers. Offline or fully air-gapped deployments require custom image builds and package mirroring.

License & commercial use

Licensed under Apache License 2.0 (OSI-approved permissive license). Permits commercial use, modification, and redistribution with proper attribution and no warranty. Source disclosure required for modifications distributed externally.

Apache 2.0 explicitly permits commercial use without royalties. However, if you modify MCPHub and redistribute it, derivative works must include license/attribution and source disclosure. Internal-only modified use carries no redistribution obligation. Recommend legal review if bundling as a proprietary product.

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	Strong
Assessment confidence	High

Security considerations

MCP endpoints require authentication by default (Bearer token or OAuth 2.0) to prevent accidental exposure; can be disabled for local/trusted environments. Dashboard login uses JWT + bcrypt. Admin password generation and change on first login documented. No security audit, penetration test results, or vulnerability disclosure policy mentioned in data. Bearer authentication, OAuth client/server modes, and database password handling should be reviewed before production deployment in high-security environments.

Alternatives to consider

Direct MCP client-to-server connections

Simpler for single-server scenarios; no orchestration layer overhead. Suitable for point-to-point AI client integration without centralized management or multi-server aggregation.

Custom API gateway (Kong, Traefik, AWS API Gateway + custom MCP adapters)

May offer more granular control, mature HA/failover, or vendor lock-in for teams deeply invested in specific cloud platforms. Requires custom MCP adapter development.

If building Python/JS AI applications, embedding MCP clients directly via LangChain tools or LlamaIndex integrations avoids a separate gateway service. Less suitable for multi-client orchestration.

Software development agency

Build on mcphub with DEV.co software developers

Explore MCPHub docs, run the Docker demo, or contact our team to discuss orchestration, authentication, and production deployment strategies for your AI tooling stack.

Talk to DEV.co

Related open-source tools

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

github-mcp-server mcpso mcphub.nvim

Related on DEV.co

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

MCP Servers AI Application Development API Development DevOps & Cloud Deployment Talk to DEV.co

mcphub FAQ

Can I run MCPHub without PostgreSQL?

Yes. File-based mode stores configuration in mcp_settings.json and requires no database. PostgreSQL is optional and needed only for production database backing, social login (Better Auth), and smart routing (pgvector).

Is MCPHub suitable for production with multiple concurrent clients?

Yes, with caveats: PostgreSQL backing is recommended. No built-in HA or load-balancing is mentioned; single service instance is a single point of failure. Horizontal scaling strategy not documented.

Can I modify and redistribute MCPHub as a closed-source product?

Apache 2.0 permits modification and commercial use, but derivative works distributed externally must include source code, license, and attribution. Internal-only modifications have no redistribution obligation. Legal review recommended before bundling as proprietary software.

What MCP servers are compatible?

Any MCP server that implements the Model Context Protocol SDK. Docker image variants (standard vs. full) affect runtime support for Rust, Playwright, and container-in-container. Verify your server's dependencies match the chosen image.

Custom software development services

From first prototype to production, DEV.co delivers software development services around tools like mcphub. Our software development agency staffs experienced software developers and web developers for custom software development, web development, integrations, and ongoing support across mcp servers and beyond.

Ready to centralize your MCP infrastructure?

Explore MCPHub docs, run the Docker demo, or contact our team to discuss orchestration, authentication, and production deployment strategies for your AI tooling stack.

Talk to DEV.co Schedule a Consultation