Ever wondered how coding agents generate code for you? Let's open up the black box and see exactly how these AI-powered coding assistants work by building one ourselves using C#.

In this session, Spencer breaks down the core mechanics behind many of today's popular coding agents. We'll explore how they understand code context, generate code, and even automate fixing bugs. By constructing our own simplified coding agent step-by-step, you'll gain practical insights into the inner workings and capabilities of these powerful - yet simple - AI tools.

By the end of this session, you'll have a deeper understanding of how AI coding agents function under the hood, and come away with a much broader understanding of AI tooling in general.

Presented at SDD London 2026 and was a highly rated session. Walks through all of the engineering and micro decisions that go into building a coding agent. It's as much a learning tool about building agentic systems as it is a walkthrough of the libraries available in C#.