mcp-context-forge and mcp-gateway
These are competitors offering overlapping core functionality—both serve as MCP gateways that centralize tool/resource management and expose unified endpoints—though IBM's offering appears more mature (3,393 vs 37 stars) and enterprise-focused with additional features like A2A/gRPC support and guardrails.
About mcp-context-forge
IBM/mcp-context-forge
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
Implements gRPC-to-MCP translation via server reflection and REST-to-MCP adaptation with automatic JSON Schema extraction, while providing OpenTelemetry-based observability across multiple backends (Phoenix, Jaeger, Zipkin). Runs as a native MCP server with 40+ plugins for protocol extensibility, Redis-backed caching for multi-cluster deployments, and built-in auth, rate-limiting, and retry policies across federated tool, agent, and API gateways.
About mcp-gateway
theognis1002/mcp-gateway
Model Context Protocol (MCP) Gateway & Registry - Central hub for managing tools, resources, and prompts for MCP-compatible LLMs. Translates REST APIs into MCP, builds virtual MCP servers with security and observability, and bridges multiple transports (stdio, SSE, streamable HTTP).
Built in Go with PostgreSQL and Redis backing, the gateway enforces fine-grained access control through JWT/OAuth2 authentication, RBAC policies, and content filtering—enabling secure multi-tenant namespaces where AI agents access isolated server groups. It implements automatic server discovery and health monitoring with failover capabilities, while supporting JSON-RPC 2.0, WebSocket, Server-Sent Events, and streamable HTTP transports to bridge diverse client needs with comprehensive audit logging and performance metrics.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work