A ChatGPT-Powered Tool for Automating Context-Aware Acceptance Criteria Generation for User Stories

Abstract

There has been a growing interest in using Natural Language Processing (NLP), such as OpenAI's ChatGPT for software engineering tasks, including requirements engineering (RE), software design, and software testing. This paper covers a practical implementation of a ChatGPT-powered prompt engineering framework designed to generate context-specific acceptance criteria for user stories. The tool automates each stage of the framework - preprocessing contextual information and generating the final tailored acceptance criteria. We outline the design and implementation of the tool in this paper and its effectiveness through two repositories. This work demonstrates the potential of large language models (LLMs) to reduce the manual effort involved in RE, streamline development workflows, and minimize rework-related costs in agile software projects.