Worked on enhancing workflow observability in the mastra-ai/mastra repository by improving how custom loggers are propagated to the execution engine. This feature addressed a critical gap in error tracking and logging during workflow execution, enabling more reliable debugging and incident response for complex workflows. The solution was implemented using TypeScript and full stack development practices, with a strong emphasis on testing and code quality. Collaboration with co-authors ensured robust integration of the logging enhancements, resulting in improved maintainability and reliability of workflow runs. The work focused on targeted improvements rather than broad changes, delivering measurable impact within a short timeframe.