The Azure DevOps MCP Server is Microsoft’s official open-source Model Context Protocol server, with 2,000+ stars on GitHub, that brings Azure DevOps directly to your AI agents — work items, repositories, builds and releases included.
If your team’s work lives in Azure DevOps, your AI assistant is blind without a bridge to it. This server is that bridge, built and maintained by Microsoft itself rather than a third party. Connected through the MCP protocol to an agent host such as GitHub Copilot in VS Code or another MCP client, it lets the model work with your Azure DevOps organisation in context: querying and updating work items, looking at repositories and pull requests, checking build and release status, and generally grounding its answers in the actual state of your project instead of stale assumptions. Because it speaks the standard protocol, the same server works across the growing set of MCP-compatible tools. It is MIT licensed and very actively maintained — commits land continuously, including in October 2026 — which matters for an integration that must track a living platform’s APIs.
The Azure DevOps MCP Server is free and open source under the MIT license. You will of course need your own Azure DevOps organisation, whose pricing is separate and set by Microsoft.
Engineering teams standardised on Azure DevOps who want their AI assistants to see the same work items, repos and pipelines they do — official support and active maintenance make it the low-risk way to connect the two.