Our project aims to utilize Large Language Models (LLMs) to produce code documentation that is comparative to human produced code documentation in Arabic and English. We will be utilizing an LLM-powered open-source framework for repository-level code documentation generation called Repoagent to generate the code documentation. The reason we are choosing RepoAgent is because it proved its effectiveness of generating high-quality code documentation. According to the paper, the code that was generated by RepoAgent was preferred by 91% of human evaluators
Research Question: How does the quality of code documentation generated by LLMs in arabic and english compare to human-written documentation in terms of readability, completeness, and usefulness.