Imagine an AI co-engineer that doesn’t just suggest code, but plans, builds, tests, and improves your app with you. In this session, I’ll show you exactly that with GitHub Copilot Agent Mode — and then take it further.

Building apps doesn’t stop at your IDE, so we’ll jump into GitHub.com to see how the GitHub Coding Agent tackles real-world backlog items directly from Issues — turning ideas into working features, writing tests, and creating pull requests.

You’ll get a clear look at the architecture that powers these AI agents, hands-on techniques for writing prompts that keep them productive, and a live demo where I use agents to build a small game and complete additional features — all driven by AI. By the end, you’ll know how to put autonomous development to work across your entire workflow, from code to cloud.

Key Takeaways
- See AI take the lead — Watch a Copilot agent plan, code, debug, and refactor an app in your IDE and on GitHub.com.

- Understand the tech behind it — Get a clear view of the architecture (Agent-Computer Interface, Model Context Protocol) and how the Coding Agent integrates into your GitHub workflow.

- Prompt like a pro — Learn how to craft prompts that help your AI co-engineer stay productive without micromanagement.