Happy slice of mcp pizzaMCP.pizza

mcp-foundry

MCP.Pizza Chef: microsoft-foundry

Microsoft has moved this experimental project to the cloud and states plainly that the code here will not be updated again, pointing people to the hosted Foundry MCP Server instead. What remains still browses the Azure model catalogue, fetches model details and sample code, checks quotas, creates projects and deployments, manages search indexes and documents, and scores answers with built-in evaluators. Using any of it requires an Azure subscription. Setup is a short settings block, but the tool assumes real Azure familiarity.

Deprecated · retired by Microsoft; use the hosted Foundry MCP Server instead
Coding
Data

Use This MCP server To

Browse the Azure model catalogue without leaving my editor Compare two models before committing to one Check which models I have quota for in a region Search my own documents through an Azure search index Score an assistant's answers with a chosen evaluator Get sample code for a model I am considering

README

MCP Server that interacts with Azure AI Foundry (experimental)

A Model Context Protocol server for Azure AI Foundry, providing a unified set of tools for models, knowledge, evaluation, and more.

GitHub watchers GitHub forks GitHub stars

Azure AI Community Discord

Important notice

MCP Server for Azure AI Foundry (experimental) has moved to the cloud, now as Foundry MCP Server (preview). Please check the official public documentation at Get started with Foundry MCP Server (preview). This GitHub repository contains the outdated information on the previous experimental server implementation, and will not be updated further. Instead, refer to the new, remotely hosted and managed Foundry MCP Server, that has the following benefits:

  • Cloud-hosted interface for AI tool orchestration: Foundry MCP Server (preview) provides a secure, scalable endpoint for MCP-compliant clients. You don't need to deploy infrastructure, enabling seamless integration and multi-agent scenarios.
  • Identity and access control: The server enforces authentication and authorization with Microsoft Entra ID. It performs all operations within the authenticated user's permissions (On-Behalf-Of flow).
  • Scenario-focused, extensible tools: The MCP Server exposes a growing set of tools for read and write operations on models, deployments, evaluations, and agents in Microsoft Foundry. The tools are extensible, letting developers and agents interact with services without knowing backend APIs or data schemas.
  • Accelerated agent and developer productivity: Natural language workflows (via MCP clients and large language models) enable rapid tool discovery and invocation, streamlining development and multi-agent orchestration.

The rest of this repo (including codebase) explains the now-deprecated MCP Server that interactes with Azure AI Foundry (experimental).

Available Tools

Capabilities: Models

mcp-foundry FAQ

Is this still maintained?
No. Microsoft moved it to a cloud-hosted service and says this version will not be updated further. The replacement it names is the Foundry MCP Server preview, documented on Microsoft Learn.
Should I use this or the hosted replacement?
The hosted one, in almost every case. It needs nothing installed, signs you in with your Microsoft account, and keeps getting new features.
Do I need an Azure account?
Yes. Everything it does runs against your own Azure subscription, and the Azure usage itself is billed to you.
Can I use this to deploy a model without opening the Azure portal?
Yes, that was one of its real strengths. It can create a services account, check quota, and deploy a model straight from a chat.
Which apps does it work in?
VS Code is the one documented, and anything else that accepts a standard MCP configuration should also start it.
How hard is the setup?
Pasting a short settings block, plus signing in to Azure and having the uv runner installed.