Welcome to Claude Code: Essentials.
As you probably know, we’re always committed to creating practical and hands-on materials. We rarely delve too much into theory or explain concepts painstakingly. We love learning by building things ourselves, so we teach that way, and many of you prefer learning this way too.
This course is no different. From the first module, we work on a real .NET solution, and we use Claude Code to change it, test it, and ship it.
What You Will Learn
We are going to take one .NET solution, a small application called Repair Desk, and work on it with Claude Code from start to finish. Along the way we will:
- Install Claude Code, sign in, and make our first change with a test to prove it works.
- Learn how the agent works, so we can tell when it is doing well and when it is guessing.
- Pick the right model and the right amount of effort for a task.
- Build a complete feature, with a database migration, across three projects.
- Write the instructions and permissions that make the agent behave the way we want.
- Set up verification we can trust, so “done” means the build and the tests say so.
- Refactor legacy code without breaking it.
- Run migrations, open pull requests, and review code the agent wrote.
- Diagnose a broken setup when things go wrong.
- Ship one feature on our own, end to end, as the final project.
Every lesson adds one thing you can do that you could not do before. By the end, Claude Code will be a normal part of how you work on .NET code.
Who This Course Is For
This course is for .NET developers who want to work with AI (artificial intelligence) tools. Coding agents are becoming a normal part of a developer’s day.
The developers who know how to direct one, check its work, and keep it inside safe limits are the ones who get the interesting work and the better offers. That is what we teach here, with Claude Code, on real .NET code.
We assume you know C# and ASP.NET Core, you have used Entity Framework Core, and you are comfortable in a terminal. We do not assume you have ever used Claude Code, or any AI coding tool.
If you have tried it once and it wrote code that looked right but did not work, you are in exactly the right place. Most of this course is about making sure that never happens to you again.
By the end, you will be able to hand a feature, a bug or a refactoring to Claude Code and know what it did. You will also prove that it works, in less time than doing it by hand. That is a skill your team will notice.
How We Teach
Every time we ask Claude Code to do something, we show you the exact prompt we typed, and right below it, the exact answer we got. Nothing is made up and nothing is retyped.
When the agent got something wrong on our machine, we leave the wrong answer on the page and show you how we corrected it. That is what working with an agent looks like.
Our recommendation is to follow along and run everything yourself. Read the lesson, then open your terminal and do the same thing.
Your results will not be word for word the same as ours, and that is normal. What matters is what the agent did, and whether the build and the tests agree with it.
Following along takes time, and it is better to know how much before you start.
Plan for 25 to 40 hours: about 5 to 10 hours of reading, the rest hands-on. Modules 0 to 3 take about an evening each. From module 4 on, plan two or three evenings per module. Module 11 takes about an hour.
Claude Code changes quickly, and so does this course. We wrote it against Claude Code v2.1.277 in September 2026 and updated it in October 2026. Lesson 0.6 lists every update, with its date.
In the next lesson, we check the four things you need before you start. And if you get stuck or need any help, do let us know, and we’ll get back to you very quickly.
Happy coding!