During October 2025, Agui worked on the ISISNeutronMuon/MDANSE repository, focusing on improving resource management and code maintainability. They introduced environment variable-based threading limits for OpenMP, BLAS, and MKL, allowing the application to control CPU usage dynamically and ensuring stable performance in multi-core environments. Using Python, Agui also refactored the geometry module by removing the unused and inaccurate random_points_on_disk function, simplifying the codebase and reducing potential errors. Their work emphasized code refactoring, environment configuration, and performance optimization, resulting in a more predictable and maintainable system that is better prepared for future enhancements and production workloads.