News
Microsoft Publishes Model Context Protocol Lab for Copilot Studio
A new GitHub repo serves as a lab for creating a Model Context Protocol (MCP) Server and using it in Microsoft Copilot Studio.
The MCP, originated by AI leader Anthropic and taken open source, is a standard that allows AI assistants to dynamically discover, access, and invoke external tools, data, and functions at runtime. Organizations can create specific MCP servers to dynamically expose their internal tools, data, and workflows to AI copilots while maintaining control, security, and governance. The new lab shows how in the Microsoft-centric space.
Copilot Studio, Microsoft's low-code tool for building, customizing, and deploying AI-powered agents that automate tasks across apps and workflows,
received MCP tech last month, with the company saying it's used to:
- Easily connect to data sources: Whether you have a custom internal API or external data providers, the MCP protocol enables smooth and reliable integration into Copilot Studio.
- Access the marketplace of existing servers: In addition to custom connectors and integrations, users can now tap into a growing library of pre-built, MCP-enabled connectors available in the marketplace. This capability gives you more ways to connect with other tools and makes using them faster and easier.
- Flexible action capabilities: MCP servers can dynamically provide tools and data to agents. This enables greater flexibility while reducing maintenance and integration costs.
It's integrated with the Power Platform and enables both business users and professional developers to create agents that work as standalone copilots, inside Power Platform apps, or embedded in other applications like Microsoft Teams or websites.
Microsoft has been busy upgrading the tool, just recently announcing an early access research preview of "computer use" for Copilot Studio wherein AI agents visually interact with any app or website -- clicking, typing, and navigating like a human (see "You Can Now Apply for Early-Stage AI Agent 'Computer Use' in Copilot Studio.") At about the same time, the company announced a new C# SDK for the MCP (see "Trending Model Context Protocol for AI Agents Gets C# SDK").
With the company embracing the MCP on a number of fronts and touting Copilot Studio for the red-hot field of agentic AI development, the new GitHub repo is a natural next step. The repo, Microsoft Copilot Studio ❤️ MCP, includes a number of examples and sample code for developers to work with. Specifically, they can create, deploy, and connect an MCP Server to Microsoft Copilot Studio using the Power Platform connector framework.
That existing Power Platform connector framework enables secure, reusable integrations between Microsoft Power Platform apps and external services through standardized, governed APIs.
Microsoft's April 28 announcement said the two technologies work together for a better experience by combining the strengths of both approaches. Traditional connectors provide secure, governed access to known services, while MCP adds dynamic, runtime extensibility by allowing copilots to discover and use new tools on demand. By wrapping MCP servers inside the familiar connector framework, developers can extend Copilot Studio's capabilities without sacrificing security, compliance, or control -- delivering a more flexible and powerful AI experience within trusted enterprise boundaries.
Connectors Are Key
Microsoft said the lab walks developers through:
- Deploying an MCP Server: Learn how to set up a server that can handle your data, models, and interactions.
- Creating a Custom Connector: Use the connector infrastructure to link your MCP server with Copilot Studio.
- Integrating with Copilot Studio: Add your MCP server as an action within Copilot Studio, enabling your agents to access the tools and data provided by the server.
Prerequisites include:
- Azure Subscription (with payment method added)
- Have a GitHub account and be logged in
- Copilot Studio trial or developer account
- Power Platform environment provisioned
What's more, Microsoft said, to be able to deploy this MCP Server and use it in Microsoft Copilot Studio using the lab, users need to create a GitHub repository from a template, deploy it as an Azure Web App, build a Power Platform connector, and add the MCP server as an action within Copilot Studio.
"By the end of the lab, you'll have a Copilot Studio agent that tells jokes by fetching them from your MCP server -- a fun and practical way to understand the integration process," the company said.
About the Author
David Ramel is an editor and writer at Converge 360.