Reproducibility is a cornerstone of good science. But documenting complex bioinformatics workflows can be a major hurdle. Researchers often struggle to create the standardized documentation needed for others to replicate their work. Enter the BioCompute Object (BCO), a standard designed to make bioinformatics workflows transparent and reproducible. But creating BCOs manually is time-consuming. A new tool leverages the power of AI to automate this process. Using Retrieval-Augmented Generation (RAG), the BCO assistant analyzes research papers and associated code, extracting the key information needed to generate BCOs. This innovative approach addresses several challenges, including the tendency of large language models (LLMs) to 'hallucinate' or invent information. The BCO assistant uses a two-pass retrieval system, refining its search to pinpoint the most relevant information from the source material. Engineered prompts for each BCO domain further improve the tool's accuracy and consistency. The tool also allows researchers to incorporate information from external sources like GitHub repositories. This simplifies the process of documenting workflows that rely on external dependencies. The BCO assistant not only saves time but also ensures compliance with the BCO standard, promoting reproducibility and collaboration in the bioinformatics community. Future development will focus on a microservices architecture for scalability and improved LLM output formatting. The aim is to refine the BCO assistant into a powerful tool that supports the continued advancement of bioinformatics research.