Worked on the microsoft/fabric-toolbox repository to implement a notebook-based table synchronization improvement for Fabric SQL endpoint metadata refresh. Refactored a Python notebook to streamline the process of refreshing tables within the Fabric environment, focusing on more accurate Lakehouse ID lookup by name and correcting the logic for SQL endpoint detection. Utilized Python and JSON to clean up code, enhance the metadata refresh workflow, and ensure reliable synchronization of table metadata. The changes reduced the risk of stale metadata in production and improved maintainability, resulting in a clearer refresh flow that supports faster onboarding and fewer regressions for developers.