The integration of Large Language Models (LLMs) in scientific research presents both opportunities and challenges for life scientists. Key challenges include ensuring transparency in AI-generated content and facilitating efficient knowledge exchange among researchers. These issues arise from the in-transparent nature of AI-driven code generation and the informal sharing of AI insights, which may hinder reproducibility and collaboration. This paper introduces git-bob, an innovative AI-assistant designed to address these challenges by fostering an interactive and transparent collaboration platform within GitHub. By enabling seamless dialogue between humans and AI, git-bob ensures that AI contributions are transparent and reproducible. Moreover, it supports collaborative knowledge exchange, enhancing the interdisciplinary dialogue necessary for cutting-edge life sciences research. The open-source nature of git-bob further promotes accessibility and customization, positioning it as a vital tool in employing LLMs responsibly and effectively within scientific communities. The manuscript is now published in Haase, R. Towards transparency and knowledge exchange in AI-assisted data analysis code generation. Nat Comput Sci (2025). https://doi.org/10.1038/s43588-025-00781-1