Jarvis Registry

A self-hosted MCP and A2A gateway with identity provider integration, access control and observability for enterprise agents

MCP serverOfficial

High risk

We rate an entry high when the tool writes to external systems, handles money, production databases or secrets, or runs arbitrary commands. The CLI installs it only with your consent.

Why this level

  • Controls agent access to internal corporate systems and data
  • Requires configuring an identity provider and stores its credentials
All reasons and checks

ascending-llc/jarvis-registry

Install

Manual install

git clone https://github.com/ascending-llc/jarvis-registry.git
cd jarvis-registry
cp .env.example .env
uv sync --all-packages
source .venv/bin/activate
docker compose --profile full up -d

Fill in .env with your identity provider credentials before starting.

This is third-party code. Review the repository files before installing.

What it does

Jarvis Registry gives you one secured entry point through which AI copilots like Cursor, Claude Desktop, GitHub Copilot and VS Code reach a company's internal tools over MCP via SSE or Streamable HTTP. It also supports registering and orchestrating autonomous agents over A2A, where an orchestrator agent coordinates worker agents through the same gateway. Access is built on OAuth 2.0 and OIDC with Keycloak, Amazon Cognito or Microsoft Entra ID integration, and an ACL engine grants permissions down to a specific tool. Semantic search over skill descriptions and tags helps agents find the right MCP server or A2A agent at runtime, alongside full request logging, OpenTelemetry tracing and Prometheus metrics.

Who it is for. For devops engineers and IT teams who need to give enterprise AI copilots governed access to internal systems.

Good fit when

  • You need one entry point for several AI copilots instead of separate integrations for each internal tool
  • You need role-based access control and a corporate identity provider instead of custom auth code
  • You need visibility into what requests agents make to internal systems and an audit trail

Not a fit when

  • You need a single local MCP server for personal use without an enterprise perimeter
  • You lack the resources to deploy and maintain an identity provider and the docker infrastructure

Example request

Deploy Jarvis Registry locally and connect Cursor to our internal MCP servers through it

Limitations

Requires configuring an identity provider through environment variables in .env and a full docker compose stack; there is no one-command cloud launch. Separate deployment guides exist for Kubernetes on AWS, Azure and GCP.

How to disable. Stop the stack with docker compose --profile full down and remove the cloned repository.

MCP

Transport
http, sse
Authentication
OAuth
Environment variables
Environment variables
identity provider credentials
required, secret
Connection details for Keycloak, Amazon Cognito or Microsoft Entra ID, set in .env before startup.

Security check

  • Controls agent access to internal corporate systems and data
  • Requires configuring an identity provider and stores its credentials

README in short

The README describes Jarvis Registry as an open-source MCP and A2A gateway with workflow orchestration from ASCENDING Inc, solving the problem of secure, governed AI agent access to internal tools. It lists an MCP reverse proxy over SSE or Streamable HTTP, integration with Cursor, Claude Desktop, GitHub Copilot and VS Code copilots, A2A agent registration and orchestration, identity management via OAuth or OIDC, tool-level ACLs, semantic skill search, and full audit trails with OpenTelemetry and Prometheus. Quick start is git clone, configure .env, uv sync, and docker compose --profile full up. Separate guides cover cloud deployment on AWS, Azure and GCP.

FAQ

Do I have to use Keycloak?

No, the README lists Keycloak, Amazon Cognito and Microsoft Entra ID as supported OAuth 2.0 or OIDC provider options.

Can I deploy this in the cloud instead of locally?

Yes, the README links to separate guides for AWS EKS, Azure AKS and GCP GKE in addition to the local docker compose setup.

Editors’ pick

An MCP server built into the Netdata agent: metrics, logs, alerts and live process, service and container data for an AI assistant

MCP serverMedium riskNo VPN needed80.7KRepository stars
Editors’ pick

GitHub's official MCP server: code, issues, pull requests, Actions and security alerts straight from the agent

MCP serverHigh risk33.3KRepository stars

Agent Skills for Google products

Agent Skills for Google products and technologies

Editors’ pick

Official Google skill collection for working with Google Cloud, BigQuery, GKE, ads and analytics from an agent

SkillHigh risk20.5KRepository stars
Official

AWS's official MCP server suite: docs, IaC, containers, serverless, databases, cost and monitoring

MCP serverHigh risk9.7KRepository stars
Foxx AIJarvis Registry

I am Foxx AI and I have already vetted this tool. Ask about install, setup or anything else, and I will keep it simple.