Microsoft Developer, joined by April, demonstrates how to create an agent using the AI Toolkit’s Agent Builder and GitHub Copilot in this Pet Planner workshop video, including integrating a custom local MCP server.

Creating an Agent with AI Toolkit and GitHub Copilot: Pet Planner Workshop Part 3

In this workshop video, April from Microsoft Developer walks through the process of building a functional agent using the AI Toolkit’s Agent Builder, with hands-on integration of a custom local MCP server. This is Part 3 of the GitHub Copilot + AI Toolkit Pet Planner workshop series.

Resources

Video Chapters

  • 00:00 - 00:02: Introduction
  • 00:03 - 03:45: Creating an AI agent with the Agent Builder
  • 03:46 - 08:37: Building a custom local MCP server for tool support
  • 08:38 - 09:56: Setting up the environment for the MCP server
  • 09:57 - 16:28: Running, testing, and integrating the MCP server with the agent

Key Topics Covered

  • Step-by-step instruction on constructing and deploying an AI agent using Microsoft’s AI Toolkit Agent Builder
  • Creating and configuring a custom local MCP (Model Communication Protocol) server to provide external tools to agents
  • Guidance on preparing the local environment for running the MCP server
  • Demonstration of agent-tool integration to expand AI-driven project capabilities
  • Insights on leveraging Microsoft Foundry’s Model and Tools ecosystem for enhanced AI development

Who Should Watch

  • Developers interested in hands-on AI agent construction and integration
  • Builders exploring GitHub Copilot and Microsoft’s AI development platforms
  • Anyone seeking to extend AI agent functionality with custom tools and local servers

For further technical deep-dives and related community discussions, reference the linked resources and join the conversation with the Microsoft Foundry community.