Programme launch offer - Save 20% Book Now

Deep Dive DD1

Test-driven Generation: A Hands-on Experience

James Lyndsay

10:30 - 12:15 CEST Wednesday 4th June

In this hands-on session, you’ll write tests that fail, and your AI will hand back code which passes those tests. You’ll go round this magic loop several times; building more tests and generating more code.

We’ll build a couple of things (a unit converter and an account handler). You can build your own tests, and we’ll have guide tests if you’d rather not. If you prefer to observe (or if we run out of bandwidth), you can look over someone’s shoulder – or watch how it works on the presenter’s machine.

Along the way, we’ll watch how the generated code grows and changes, and consider our influence over it. We’ll use coverage tools, syntax tools and change control to escape common pitfalls. We’ll explore our test-passing running system to expose the unexpected. We’ll tune our prompts to suit our purposes, and look at the scripts and tools that make this possible.

Before we go, we’ll wonder out loud whether our needs at work can be met by an ability to build things that pass our tests.

This session uses a browser and needs you to type. You’ll gain access to your own temporary browser-based development environment (Python or JavaScript): editor, file browser, command line, test runnner, coverage tool, static code checker, change control, target UI, AI wrangler and (Anthropic/Claude) API key – and a short, open-source, readable script that brings all of these together. Bring an internet-connected device that you’re happy to type on.