Visual Studio Code, with Carlotta Castelluccio, showcases how developers can use GitHub Copilot and the MCP server to programmatically interact with Microsoft Learn Docs for grounded, documentation-based AI answers.

Use Copilot and MCP to Query Microsoft Learn Docs

Presented by: Carlotta Castelluccio

Discover how to combine GitHub Copilot and the Microsoft Docs Model Context Protocol (MCP) Server tools to enhance your development workflow. This session covers both an overview of MCP and a live demo illustrating how to query Microsoft Learn Docs using a custom Python client.

What is MCP?

The Model Context Protocol (MCP) is a protocol and tooling from Microsoft designed to provide AI-powered, contextually relevant answers by grounding responses in official Microsoft Documentation. MCP integrates with the Microsoft Learn platform and facilitates prompt-driven development workflows.

Using MCP with GitHub Copilot

Setup: The video demonstrates how to set up and connect to the MCP Server, walking through the steps in Visual Studio Code.
Querying Docs: Learn how to use GitHub Copilot with your own Python client to send queries to the MCP and retrieve documentation-backed answers.
Demo: Key moments include a walkthrough of how questions are processed via Copilot and how results are sourced from Microsoft Docs.

Key Resources

Chapters

00:00 Intro
00:44 MCP Overview
02:05 Microsoft Learn MCP Server
02:40 Demo - Queries
10:10 In summary
10:50 Wrap-up

Connect with VS Code

X: https://x.com/code
Bluesky: https://bsky.app/profile/vscode.dev
YouTube: https://youtube.com/code
LinkedIn: https://www.linkedin.com/showcase/104107263
GitHub: https://github.com/microsoft/vscode

This session provides developers with practical steps and best practices for leveraging AI-powered tools in a documentation-driven context. The demo highlights prompt-driven development and the ability to ground AI responses directly on product documentation for improved accuracy and productivity.