Worked on the llamastack/llama-stack repository to improve the reliability and clarity of the Retrieval-Augmented Generation (RAG) tutorial. Addressed a recurring issue where incorrect handling of server-assigned vector_db_id values led to 400 Bad Request errors by updating the tutorial to capture the ID from the registration response, ensuring alignment with current API behavior. Focused on documentation and tutorial development using Markdown, the changes streamlined onboarding for new users and reduced support friction. The work demonstrated attention to API integration details and response handling, resulting in a smoother developer experience and more accurate, up-to-date instructional materials for the project.