GH Repository · 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.
- stars
- 4,475
- 30-day movement
- +279/day
- Related entries
- 60
- Connections
- 4
rustpython/uvmcp-serverPythonpythonprompt-engineeringnodemcpllm-agentsagentsaiapi-gatewayobservabilitykubernetestoolsmodel-context-protocolauthentication-middlewaremakedockerfastapidevopsfederationgenerative-aigateway
IBM's mcp-context-forge is an AI gateway, registry, and proxy that fronts MCP, A2A, and REST/gRPC APIs through a single unified endpoint. It adds centralized discovery, guardrails, management, optimized agent and tool calling, and plugin support, implemented in Python with FastAPI.
Reach for it when you need one managed gateway layer for discovery, authentication, guardrails, and observability across mixed MCP, A2A, and REST/gRPC backends.
Use it to
- Expose multiple MCP servers through a single endpoint
- Apply guardrails and JWT authentication to agent API calls
- Centralize tool and API discovery in a registry
- Deploy an AI gateway on Kubernetes with Docker
- Observe and manage agent tool-calling traffic
For Teams building and operating agent or MCP infrastructure
- Role
- mcp-server
- Language
- Python
- Licence
- Apache-2.0
- Forks
- 865
- Open issues
- 701
- Last push
- 2026-09-15
- Latest release
- v0.1.0 · 2025-06-05
topicsai-gatewaymcpproxyguardrailsobservabilityfederation