ContextForge review

Q: How do I install ContextForge?

Run pip install mcp-contextforge-gateway. The mcp-contextforge-gateway package needs Python 3.12 or newer, below 3.14. The install section above has the other options from the project's README.

ContextForge is an open source gateway, registry, and proxy that federates MCP, A2A, and REST/gRPC APIs behind one endpoint.

Updated Nov 1, 2026. Review written Oct 22, 2026 from the project's own sources; usage data refreshed every Monday.

TL;DR

ContextForge is an open source registry and proxy from IBM that federates MCP servers, A2A agents, and REST/gRPC APIs into one endpoint for AI clients. It suits teams that need centralized governance, discovery, and observability across AI infrastructure. It can wrap non-MCP services as virtual MCP servers and ships with an Admin UI and OpenTelemetry tracing.

4.7kGitHub stars
+46new stars, week of Oct 25
8.5kdownloads a week, +2% vs 4-week avg
#47in MCP and agent tools this week

ContextForge at a glance

Category: MCP and agent tools
Maintained by: International Business Machines (organization, @IBM on GitHub)
License: Apache 2.0
Language: Python, HTML, JavaScript
Repository: github.com/IBM/mcp-context-forge
Website: ibm.github.io/mcp-context-forge
PyPI package: mcp-contextforge-gateway 1.1.0
Latest release: v1.1.0 on GitHub, Oct 20, 2026
First published: May 8, 2025

What ContextForge does

ContextForge federates multiple MCP servers and REST services and exposes them through a single, unified interface.
ContextForge wraps REST and gRPC services as virtual MCP servers, using server reflection for gRPC-to-MCP translation.
ContextForge routes A2A agents, including OpenAI-compatible and Anthropic agents, through its Agent Gateway.
ContextForge provides built-in auth, retries, and rate limiting, with Basic, JWT, OAuth, and inbound mTLS options.
ContextForge keeps registries for prompts (Jinja2 templates with versioning), resources, and tools with input validation.
ContextForge traces tools, prompts, and gateway operations with OpenTelemetry to backends like Phoenix, Jaeger, and Zipkin.
ContextForge includes an Admin UI for real-time management, configuration, and log monitoring, plus a separate Web UI container.

How it works

ContextForge runs as a fully compliant MCP server, installed from PyPI or run as a Docker or Podman container. You register upstream MCP servers or APIs, then bundle their tools into virtual servers. AI clients connect over HTTP, JSON-RPC, WebSocket, SSE, or Streamable HTTP. Larger deployments run on Kubernetes with Redis-backed federation and caching, and Docker Compose sets up PostgreSQL and Redis.

How to install and start using ContextForge

With pip

pip install mcp-contextforge-gateway

With uv

uv pip install mcp-contextforge-gateway

With Docker

docker pull ghcr.io/ibm/mcp-context-forge:latest

Requirements

Python 3.12 or 3.13 for version 1.1.0 on PyPI (requires_python <3.14, >=3.12).
JWT_SECRET_KEY and AUTH_ENCRYPTION_SECRET must be set in every environment, including local development.
curl and jq for the smoke-test step only.

First steps

Generate the required secrets with the bundled init script before the first run.
Start the gateway with mcpgateway --host 0.0.0.0 --port 4444.
Create a bearer token with mcpgateway.utils.create_jwt_token and call the /version endpoint to smoke-test.
Register an MCP server through the /gateways endpoint, then create a virtual server bundling its tools.

Who ContextForge is for

Platform teams putting many MCP servers and agents behind one governed endpoint.
Developers exposing existing REST or gRPC services as MCP tools without rewriting them.
Teams that need tracing, auth, and rate limiting for agent and tool calls.
Organizations deploying on Kubernetes across multiple clusters.

Limitations to know

ContextForge container images do not support arm64 in production; Apple Silicon users can use Rosetta or install from PyPI.
ContextForge will not start with placeholder secret values.
ContextForge local Docker image builds need a wheel closure produced only by CI, so pull the pre-built image instead.

Is ContextForge growing? Usage data

In the week of Oct 25–31, 2026, ContextForge ranked #294 on the Devtoolwatch Board and #47 in MCP and agent tools, with a momentum score of 32 out of 100. Its main package had 8.5k downloads that week, +2% against its own 4-week average.

New GitHub stars per week for the last 12 weeks, from GitHub's stargazer history (today's stargazers by the week they starred).

Show the weekly numbers

New GitHub stars per week for ContextForge
Week of	New stars
Aug 9, 2026	47
Aug 16, 2026	27
Aug 23, 2026	37
Aug 30, 2026	43
Sep 6, 2026	41
Sep 13, 2026	36
Sep 20, 2026	31
Sep 27, 2026	35
Oct 4, 2026	32
Oct 11, 2026	53
Oct 18, 2026	42
Oct 25, 2026	46

On Hacker News

IBM Type, 326 points and 121 comments (Nov 8, 2017)
MCP Gateway and Registry, 73 points and 53 comments (Aug 25, 2025)

The Board ranks tools by four signals each week: new stars, new stars for the repo's size, added downloads and download growth beyond the market. How the ranking works.

ContextForge alternatives

Other MCP servers and agent tools on the Board in the week of Oct 25–31, 2026, by momentum:

MCP and agent tools ranked by momentum, the week of Oct 25–31, 2026
Rank	Tool	Stars	Downloads/wk
#1	REAREA is an MCP server and CLI that lets coding agents reverse engineer native binaries, JavaScript and Electron apps, .NET assemblies,…	415.5k	760.6k
#2	ha-mcpha-mcp is an unofficial Model Context Protocol server that lets AI assistants control and manage Home Assistant using natural language.	5.4k	118.1k
#3	TregTreg is a registry and proxy that gives AI agents one token for a catalog of paid third-party API endpoints and team-shared tools.	6.9k	6.7k
#4	Chrome DevTools MCPChrome DevTools MCP is an MCP server and CLI that lets coding agents control and inspect a live Chrome browser using Chrome DevTools.	54.6k	2.7M
#5	codebase-memory-mcpHigh-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 162…	47.9k	–
#6	open-computer-use👾 Open Computer Use – Open-Source Alternative to Codex Computer Use	2.8k	–

All MCP servers and agent tools, ranked

Recent ContextForge releases

v1.1.0, Oct 20, 2026
v1.0.11, Sep 28, 2026
v1.0.7-20260921, Sep 21, 2026
v1.0.10, Sep 7, 2026
v1.0.9, Sep 1, 2026

ContextForge FAQ

Is ContextForge free and open source?

Yes. ContextForge's source code is on GitHub (IBM/mcp-context-forge) under the Apache 2.0 license, so you can use it for free.

How do I install ContextForge?

Run pip install mcp-contextforge-gateway. The mcp-contextforge-gateway package needs Python 3.12 or newer, below 3.14. The install section above has the other options from the project's README.

Is ContextForge still maintained?

Yes. The IBM/mcp-context-forge repository was last updated on Oct 21, 2026, and the latest release, v1.1.0 on GitHub, came out on Oct 20, 2026.

How popular is ContextForge?

ContextForge has 4.7k GitHub stars as of Nov 1, 2026. It gained 46 new stars in the week of Oct 25–31, 2026, against a 12-week average of 39 a week. Its main package had 8.5k downloads that week, +2% against its own 4-week average. That made it #47 among MCP servers and agent tools on the Devtoolwatch Board.

What are the best alternatives to ContextForge?

Other MCP servers and agent tools on the Devtoolwatch Board in the week of Oct 25–31, 2026, by momentum: REA (#1), ha-mcp (#2), Treg (#3), Chrome DevTools MCP (#4), codebase-memory-mcp (#5). The alternatives table above has their numbers.

What is ContextForge?

ContextForge is an open source registry and proxy that federates MCP servers, A2A agents, and REST/gRPC APIs into one endpoint for AI clients. It adds centralized governance, discovery, and observability, and supports plugins.

Can ContextForge turn a REST or gRPC API into MCP tools?

Yes. ContextForge wraps non-MCP services as virtual MCP servers. REST APIs are adapted with automatic JSON Schema extraction and auth support, and gRPC services are translated through server reflection.

Which transports does ContextForge support?

ContextForge supports HTTP, JSON-RPC, WebSocket, SSE with configurable keepalive, and Streamable HTTP. stdio transport is available for server-side use, and mcpgateway.translate can expose a stdio server over SSE and Streamable HTTP.

What changed in ContextForge 1.1.0?

ContextForge 1.1.0 turns modern MCP protocol negotiation on by default. It adds federated completions, MCP catalog pagination, Origin/Host transport security enforcement, and tool-name collision detection. It also removes the in-tree mcp-servers/ directory.

Which observability backends does ContextForge work with?

ContextForge uses OpenTelemetry (OTLP) and supports Phoenix, Jaeger, Zipkin, Tempo, DataDog, and New Relic. It records LLM metrics such as token usage and costs, and it adds no overhead when tracing is disabled.

Sources

ContextForge README badge

Maintainers can show the tool's weekly Board rank in their README:

[![Devtoolwatch](https://index.devtoolwatch.com/badge/ibm-mcp-context-forge.svg)](https://index.devtoolwatch.com/tools/ibm-mcp-context-forge/)

How this page is made: Devtoolwatch writes each review from the project's README, releases, package metadata and website, with an AI model, then checks every claim against the source it cites and drops what the sources don't support. The numbers come from Devtoolwatch's daily GitHub snapshots, GitHub's stargazer history, and the npm and PyPI download APIs. Spotted an error? Email [email protected].