Developed targeted regression tests for the Telemetry Restore Metrics feature in the All-Hands-AI/agent-sdk repository, focusing on validating cost tracking and metrics restoration within the telemetry system. Leveraged Python to expand test coverage, ensuring that production metrics remain reliable and reducing the risk of regressions affecting cost monitoring. The approach centered on comprehensive testing strategies, integrating metrics tracking and telemetry validation to support data-driven decision-making. By enhancing the robustness of the restore_metrics functionality, the work contributed to more dependable telemetry operations. Emphasis was placed on thorough testing and quality assurance, utilizing Python and established testing methodologies throughout the development process.