Worked on enhancing autograding reliability for the PrairieLearn/PrairieLearn repository by addressing a key issue in Python-based notebook execution. Focused on improving traceback handling within the Python Autograder, the developer implemented a targeted patch to line caching for Jupyter Notebook (ipynb) files, ensuring that error tracebacks now accurately reference the correct source lines in student notebooks. This backend development effort, using Python and autograding technologies, aimed to reduce confusion during debugging and streamline support workflows. Additional targeted tests were added to safeguard against regressions, resulting in a more robust autograding experience for both students and support teams.