AI-Powered Multi-Agent Framework for Automated Unit Test Case Generation: Enhancing Software Quality through LLM’s

Recent years have witnessed an enormous rise in the design, repair and the enhancement of software automation tests. The reliability of program’s unit testing has major impact on its overall performance. The anticipated influence of Artificial Intelligence advancements on test automation methodologies are significant. Many studies on automated testing implicitly assume that the test results are deterministic, means that similar tests faults remain same. The precision of software is largely ensured by unit testing. But writing unit tests manually is a time-consuming process, which leads us to drive into "Automation Analysis". Recent years comprised the application of Large Language Models (LLM’s) in numerous fields related to software development, especially the automated creation of unit testing.However, these frameworks require more instructions, or few shot learnings on sample tests that already exist. This research provides a comprehensive empirical assessment of the efficiency of LLM’s for automating unit testing production, with no need for further manual analysis. The method we employ is put into practice for test cases, an adaptable Agents and LLM-based testing framework that evaluates test cases generated, by reviewing and re-writing them in different phases. Evaluation of this test cases was done by using mistral-large LLM Model. The analysis results that developed acquired an overall coverage of 100% for code given. Finally, to enhance the typical evaluation, this research suggests and concludes that LLMs, can be successfully incorporated into present practices, through adaptative instructions and improvements.