Hugues developed and launched an educational MCP server for the punkpeye/awesome-mcp-servers repository, focusing on math operations, statistics, visualization, and persistent workspaces. The project emphasized data science and educational tools, with all documentation authored in Markdown to ensure clarity and accessibility. By introducing persistent compute environments, Hugues enabled reproducible experiments and streamlined onboarding for learners and researchers, reducing setup time and supporting long-running analytics tasks. The work included targeted updates to the Data Science Tools documentation, reflecting the new server’s capabilities. This release laid a solid foundation for scalable, training-focused data science education within the platform’s existing infrastructure.