Worked on the DataDog/datadog-agent repository to address telemetry accuracy during cluster leadership transitions. Developed a cleanup mechanism in Go within the cluster store to reset danglingConfig and unscheduledCheck metrics whenever the agent switched between leader and follower roles. This targeted bug fix included comprehensive unit testing to ensure that gauge metrics were reliably cleared, preventing stale telemetry data and reducing false alerts during failovers. The work focused on backend development and telemetry, enhancing the reliability and observability of the cluster-agent. The solution was isolated, well-tested, and delivered measurable improvements to cluster stability and monitoring accuracy in production environments.