Get started with the Azure MCP Server

The Azure MCP Server enables AI-powered development tools to interact with Azure cloud services through the Model Context Protocol (MCP). It provides a unified way to manage Azure resources, deploy applications, and query cloud services directly from your development environment.

Connect to Azure MCP Server using various tools, languages, and frameworks. Use it to manage Azure resources through natural language conversations, build automation scripts, or integrate Azure operations into your applications.

Explore and contribute to the Azure MCP Server on GitHub.

Install Azure MCP Server using a package manager

You can optionally install Azure MCP Server directly using one of the supported package managers. Package manager installation offers several advantages, including centralized dependency management, CI/CD integration, support for headless environments, version control, and project portability.

Note

You don't need to install the Azure MCP Server to use it. The approaches explained in the Connect to Azure MCP Server section either include Azure MCP Server as part of their installation, or run Azure MCP Server directly via package manager commands.

For more information, see Install with a package manager.

• NuGet: Azure.Mcp
• NPM: @azure/mcp
• PyPI: msmcp-azure

Connect to Azure MCP Server

You can connect to Azure MCP Server using the following tools and programming frameworks.

Use a code editor

Azure MCP Server works with AI-powered code editors and tools that support the Model Context Protocol. Learn how to get started:

• Cline
• Cursor
• IntelliJ
• Visual Studio
• Visual Studio Code
• Windsurf

Use GitHub Copilot tools

Connect to Azure MCP Server using other tools and services, such as GitHub Copilot coding agent and Docker. Learn how to get started:

• GitHub Copilot CLI
• GitHub Copilot SDK
• GitHub Copilot coding agent

Use a Docker container

For easy deployment and isolation, you can run the Azure MCP server as a Docker container.

Use programming languages & frameworks

Connect to Azure MCP Server using programming languages and frameworks. This documentation currently provides guidance and examples for Python and .NET.

Python

Get started with Azure MCP Server and Python to enhance your apps and workflows.

• Use Python MCP client libraries to connect directly to Azure MCP Server.
• Build automation scripts that manage Azure resources.
• Integrate into web frameworks like Django, Flask, or FastAPI.
• Incorporate Azure operations into data science workflows with Jupyter notebooks.

.NET

Get started with Azure MCP Server and .NET to enhance your apps and workflows.

• Create console applications and command-line tools for Azure management.
• Build ASP.NET Core web applications with integrated Azure capabilities.
• Develop Azure Functions that leverage Azure MCP Server for resource management.
• Create Windows desktop applications with Azure integration.

The Azure MCP Server provides flexibility to work with Azure in the way that best fits your development style and requirements, whether through interactive AI-powered editors or programmatic integration in your applications.