Summary
Unit 6 of 6 · Module 13: Develop an AI agent with Microsoft Agent Framework
Read this lesson
Snapshots of the module
In this module, you met Microsoft Agent Framework, the next generation of Semantic Kernel and AutoGen. It gives you building blocks that fit together: model clients, sessions, context providers, function tools, MCP clients, middleware, and workflows. Every agent shares one base class, so switching providers means changing only the client, and Foundry Agent Service is the recommended provider because it keeps conversation history on the service side. You created a Foundry chat client with your endpoint, model, and credential, built an agent with instructions and tools, and ran it with or without streaming. You added hosted tools and your own functions, described with Annotated fields or the tool decorator, and saw approval mode and agents used as tools.
What's next
To go further, read the Agent Framework overview and its guides to agents, tools, and sessions. Next, in Module 14, you'll put several agents together and orchestrate them as a team.
Facts from Microsoft's documentation: Microsoft Agent Framework Overview (CC BY 4.0), rewritten in Loop's own words and examples. Narration, animation, and quiz by Loop. Microsoft, Azure, and Microsoft Foundry are trademarks of Microsoft. Loop is independent and not affiliated with or endorsed by Microsoft.