Microsoft has made waves in the developer community by releasing the Azure DevOps Model Context Provider (MCP) Server in public preview. This innovative tool bridges the gap between GitHub Copilot and Azure DevOps, allowing developers to interact with their project data effortlessly while immersed in their coding environment, such as Visual Studio Code (VS Code).
At its core, the MCP Server acts as a local intermediary that connects GitHub Copilot in Agent Mode with a developer’s Azure DevOps instance. By leveraging natural language prompts, developers can now access structured project data, including work items, pull requests, test plans, builds, and wiki entries, right from their IDE. As Microsoft’s product manager Dan Hellem points out, this empowers the assistant to deliver answers that are not just general, but specifically tailored to the developer’s ongoing projects.
One of the most compelling features of the MCP Server is its design as a local service. By running locally, it ensures that sensitive project data remains within the developer’s network, providing peace of mind about security and privacy. Installation is straightforward, requiring Node.js and integration with GitHub Copilot through a simple configuration file and Azure CLI authentication.
This initial version of the MCP Server focuses on essential functionalities within Azure DevOps. From the ability to list projects, repositories, and build pipelines to creating and editing work items using natural language, the experience is intuitive and efficient. GitHub Copilot not only assists in suggesting content but also simplifies the process of generating test cases and linking pull requests to backlog items. Lyon Till, a Senior Software Engineer at Microsoft, emphasizes that this eliminates the frustrating context-switching developers often face, streamlining the entire DevOps workflow within VS Code or Visual Studio.
To explore this innovative tool, developers need to follow a few simple steps: install the MCP Server from the official GitHub repository, sign in using the Azure CLI, configure the local mcp.json file, and start the service. Once set up, developers can issue commands like “list work items assigned to me” or “create a user story for login failure handling.” The MCP Server translates these conversational commands into authenticated REST API calls to Azure DevOps, making collaboration and task management more fluid.
It’s important to note that this preview release currently supports Azure DevOps Services and does not extend to on-premises Azure DevOps Server. Microsoft is actively seeking feedback from early adopters via GitHub Issues, encouraging input that could shape future enhancements and features.
In addition to the Azure DevOps MCP Server, Microsoft launched a separate MCP Server for Azure resource queries earlier this year. This tool allows developers to retrieve information about various Azure services—including Key Vault, Cosmos DB, and Storage Accounts—using natural language semantics. The Azure DevOps MCP Server builds upon this similar framework, enhancing the developer experience in managing cloud resources and project data alike.
For those interested in getting started, detailed documentation, setup guidance, and the source code are readily available on GitHub. The preview is open to developers using Azure DevOps Services and GitHub Copilot Business or Enterprise licenses, opening doors to streamlined workflows for teams looking to elevate their productivity.
Inspired by: Source
