Generating CS1 Coding Questions using OpenAI

In CS1, to assess student knowledge, instructors prepare exam questions that often include code snippets. Due to the significant amount of time and effort required to create high-quality exam questions, instructors often only produce a single version of the exam. This results in all students receiving the same set of questions, which raises the possibility of plagiarism. In this paper, we propose a tool that allows computing science educators to generate a number of variations of a given code snippet, where the pedagogical intent of the code remains the same, but the code is mutated.