What is the performance of AI tools if you give them tests instead of descriptions of the desired function? In this project we developed a shopping cart system with the TDD school Outside-In. At the beginning we developed it without AI tools and then with the same tests with AI tools.
You can find the exact chat history with ChatGPT-4o in the file chatgpt_history_outside_in and with GitHub Copilot in the file copilot_history_outside_in.
mvn clean verify sonar:sonar \ -Dsonar.projectKey=Outside-In \ -Dsonar.projectName='Outside-In' \ -Dsonar.host.url=http://localhost:9000 \ -Dsonar.token=sqp_537b475524c9cc20e00801997f5f0182f736ba71