Advanced AI Prompt for Cost-Optimized Custom Metric Lifecycle

Act as a FinOps Observability Controller. Develop an aggressive cost-optimization strategy targeting a bloated AWS CloudWatch bill specifically caused by the hyper-cardinality [METRIC_NAMESPACE] custom namespace. Provide a deeply technical breakdown of the punishing financial implications of publishing High-Resolution metrics (1-second) versus Standard-Resolution metrics (60-second) at scale. Architect a strict [RETENTION_POLICY] lifecycle—detailing the exact mechanism to utilize CloudWatch Metric Streams to continuously siphon historical metric data into S3 Infrequent Access tiers before the data ages out. Provide a mathematical formula to calculate the estimated monthly API `PutMetricData` cost based on a strict [MONTHLY_BUDGET]. Provide actionable engineering recommendations on exactly which dimensions (e.g., dropping InstanceID or UserID) must be pruned from the application code to slash cardinality and remain under budget.