Session 3 - Logic, Tests

In this session we're going to have Gemini help us build the logic for our application, complete with tests. As always, if you prefer to use Copilot with a different model, do it! Your results might be slightly different, which means you get to figure out why and adjust your prompt as needed. This is a good thing - it's how we learn.

Our tasks in this session are:

• Formalizing our spec with a /docs directory and using it as context when we need to, rather than having it live in our instructions.
• See what it's like to do test-first development with AI, having AI take a stab at writing the initial tests for us.
• Fill out the logic, making sure the tests pass.
• Perform a code review to lower the chance of crap code.

Off with us!