An LLM-Powered Tool for Enhancing Scientific Open-Source Repositories

Nikolay Nikitin, Andrey Getmanov, Zakhar Popov, Ulyanova Ekaterina Alekseevna, Yaroslav Aksenkin, Ilya Sokolov, Alexander Boukhanovsky

Keywords: open-source, code analysis, scientific software, repository improvement, README

TL;DR: Tool that makes scientific open source projects better using LLM agents

Abstract: We present OSA (Open Source Advisor), an open-source tool that uses large language models (LLMs) to improve scientific repositories. OSA uses an agent-based architecture to automate tasks such as generating README files, docstrings, and documentation, configuring CI/CD pipelines, creating tag annotations, and analyzing repositories to provide recommendations on best practice. Unlike code-generation tools, OSA focuses on enhancing existing repositories. We have validated OSA's effectiveness across diverse scientific repositories in multiple domains. OSA is available in https://github.com/aimclub/OSA.

Submission Number: 17