mcp-hub and metamcp
One tool serves as a comprehensive middleware and gateway for Model Context Protocol (MCP) interactions, while the other is a centralized manager for MCP servers, making them complementary tools that can be used together to manage and orchestrate an MCP ecosystem.
About mcp-hub
ravitemer/mcp-hub
A centralized manager for Model Context Protocol (MCP) servers with dynamic server management and monitoring
Provides a unified HTTP endpoint that aggregates multiple MCP servers with automatic capability namespacing, allowing any MCP client to access all managed servers without individual configuration. Supports local STDIO servers and remote endpoints via streamable-http/SSE transports with OAuth 2.0 and automatic reconnection. Built-in REST API, real-time SSE events, and VS Code configuration compatibility enable dynamic server management, hot reload, and environment variable interpolation.
About metamcp
metatool-ai/metamcp
MCP Aggregator, Orchestrator, Middleware, Gateway in one docker
Supports dynamic MCP server aggregation with middleware injection and tool-level filtering, enabling selective exposure of capabilities across namespaces. Implements a proxy architecture using stdio transport with environment variable management, allowing configuration-driven composition of multiple MCP servers into unified endpoints with granular access control via API-key or OAuth authentication.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work