Agentic Coding

Start Here

New here? Read this first. It is a short orientation: what this course is, how it is put together, and where to begin.

What this is

A hands-on course on building real software with AI coding agents, in fifteen chapters over three levels. The material comes from real development work rather than hypotheticals: the examples, the numbers and the failures are drawn from shipping projects, and each insight cites the commits behind it.

How it is put together

The chapters are numbered by level, the way undergraduate courses are. The first digit is the level: 1xx is one human with one agent, 2xx is a practitioner's method at project scale, and 3xx is one human running a fleet of agents.

Each chapter is one teaching unit in three parts: a theory page that works through the chapter's ideas with real examples, an exercises page for hands-on practice at the keyboard, and a reference card, a one-page summary to keep beside you afterwards.

Running alongside the chapters are the Agentic Coding Insights: short, self-contained lessons, each one a single idea with the evidence behind it, and each homed in one chapter. Two case studies and a reference section round out the set.

The chapters

1xx, first year

2xx, second year

3xx, third year

How to use it

There are two ways through. The intended path for a first read is sequential: start at 101 and work through the chapters in number order, since each chapter builds on the ones before it. If you are already comfortable with the basics, use the course as a reference instead, and go straight to a reference card or to an insight by topic.

Where to begin