The Azure MCP Server
One early implementation of MCP on the Microsoft’s platform is the open supply Azure MCP Server, which not too long ago entered public preview. This gives a standard dealer for AI entry to key Azure companies. Like many current Azure initiatives, that is open supply, with code accessible on GitHub. It provides entry to a lot of the Azure platform, together with databases, storage, and companies just like the Azure CLI.
Help for the Azure CLI (and the Developer CLI) is attention-grabbing because it permits your MCP-powered brokers to make use of Azure straight, treating MCP calls as an operator. This allows you to construct brokers that present a pure language self-service interface to Azure, for instance, taking an outline of an infrastructure after which constructing the Arm templates wanted to deploy it. You possibly can even think about a multimodel agent that analyzes an image of a whiteboard sketch, creates an outline of the sources wanted to implement the sketch, after which deploys it so that you’re able to construct code. Different system administration companies accessible by means of the Azure MCP Server embrace itemizing present useful resource teams and utilizing KQL to question Azure Monitoring logs.
Utilizing Azure MCP server with GitHub Copilot Chat
Because it’s based mostly on MCP, this new server works from any AI device that helps MCP, for instance, the GitHub Copilot Agent Mode. Merely add the server to your tenant and begin asking questions through Copilot, both straight or with the Visible Studio Code integration. In observe, this final choice is an efficient approach to learn to use MCP and to construct the prompts in your individual MCP-based AI functions.
