Summary
Unit 7 of 7 · Module 15: Discover Azure AI Agents with A2A
Read this lesson
The island map
In this module, you connected agents that live on different servers with the A2A protocol. An agent says what it can do with skills, listed on an agent card at a well-known address, so others can discover it at run time instead of hard-coding it. An agent executor turns A2A requests into your agent's own logic, and puts messages, task updates, and artifacts on an event queue. A server, with a request handler, a task store, and a Starlette app run by Uvicorn, makes the agent reachable. And a client fetches the card and sends requests, waiting for the whole reply or streaming it, and handling either a direct message or a task. Microsoft Foundry can also expose an agent over A2A for you, and call remote A2A agents with its A2A tool.
What's next
To go further, read the A2A pages in the Agent Framework and Foundry docs. That completes the agents learning path. Next, you'll move on to building natural language solutions.
Facts from Microsoft's documentation: Enable incoming A2A on a Foundry agent (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.