Generative Chat Model for GitHub Repository Using LLMs

Authors- Dharmaraj J, Graceshon J S, Murugan S, Dr. J. Yogapriya

Abstract-GitHub repositories serve as vital resources for software development, containing source code, documentation, and configuration files that define a project’s architecture. However, navigating and understanding complex repositories can be challenging, particularly for developers unfamiliar with a given project. This paper proposes an interactive platform that leverages Natural Language Processing (NLP) and a Large Language Model (LLM) to enable users to query GitHub repositories efficiently. By inputting a repository URL, users can ask specific questions and receive detailed, context-aware responses generated by the LLM. The system processes various file types, including README.md, source code, and configuration files, to extract relevant insights. The frontend is developed using HTML, CSS, and JavaScript, while the backend utilizes Flask for data processing and routing. This model surpasses traditional keyword-based searches by offering human-like, semantically rich responses, significantly improving repository accessibility and developer productivity. The proposed system provides an innovative approach to enhancing software repository comprehension using AI-driven conversational agents.

DOI: /10.61463/ijset.vol.13.issue2.190