OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
MCP & Tooling

ContextForge

ContextForge is an open-source Python registry and proxy that federates MCP, A2A, REST and gRPC services behind a unified endpoint. It provides discovery, centralized governance and observability for agent and tool calls, with plugins and options for translating or virtualizing APIs as MCP tools.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

ContextForge — OpenViST project profile in MCP & Tooling.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use ContextForge when an agent environment needs a centrally managed registry and routing layer across multiple tools, agents or APIs, especially when those services span MCP, A2A, REST and gRPC.

Who it is for

  • Teams integrating agent clients with tools and APIs that use different protocols.
  • Platform and infrastructure engineers managing shared tool discovery, routing and controls.
  • Developers who need to expose REST or gRPC services through MCP-compatible interfaces.

Documented uses

  • Federate multiple MCP servers and REST services behind one endpoint.
  • Route calls to external A2A agents and integrate REST or gRPC APIs into agent tool workflows.
  • Virtualize legacy APIs as MCP-compliant tools or servers.
  • Apply shared authentication, retries and rate limits to routed API calls.
  • Send gateway tracing to OpenTelemetry-compatible observability backends.
Before you choose

Requirements & limits

The project is open source under the Apache-2.0 license. The README documents installation through PyPI and Docker, and requires real JWT and encryption secrets for startup.

  • ContextForge is an infrastructure gateway and registry, not an agent model or end-user chat application.
  • Its protocol scope includes MCP, A2A, REST and gRPC, with gRPC-to-MCP translation and REST-to-MCP adaptation described in the project documentation.
  • The gateway includes authentication, retries and rate limiting; teams should account for their configuration and secret-management needs.
  • The README describes a Docker Compose stack with PostgreSQL and Redis, plus Kubernetes multi-cluster federation with Redis-backed caching.
  • The documented production container path does not support arm64; the README points Apple Silicon users to Rosetta or PyPI installation.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Start ContextForge with Docker Compose

Prepare the documented Compose environment and start the gateway stack.

Before you start

  • A local Docker Compose setup.
  • A clone of the ContextForge repository.
  • A .env file populated with real JWT_SECRET_KEY and AUTH_ENCRYPTION_SECRET values.

Steps

  1. Clone the repository and enter its directory.

  2. Copy .env.example to .env, then run the documented secret initialization command to patch real secrets into the file.

  3. Pull the prebuilt gateway image from GHCR and set IMAGE_LOCAL in .env as shown in the README.

  4. Run docker compose up -d, then inspect service status with docker compose ps.

Expected outcome

The Compose stack starts with ContextForge and its supporting services; the README identifies the Admin UI at http://localhost:8080/admin.

Guide limitations

  • The gateway will not start with placeholder secrets.
  • The documented production container path does not support arm64; the README notes Rosetta or the PyPI installation path as alternatives for Apple Silicon machines.
  • This setup recipe does not configure upstream tools or verify their calls.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
9 October 2026
Original website title
IBM/mcp-context-forge · Official homepage title · retained in its source language
How we review sourcesSuggest a correction