aiops-platform-optimization-agent

What problem does it solve?

This Skill reduces platform cost and waste by continuously finding and applying safe infrastructure optimizations across Kubernetes, Kafka, and analytical databases.

Core Features & Use Cases

• Autonomous optimization loop (observe→analyze→recommend→apply): Detects optimization opportunities, prioritizes them, and applies low-risk changes.
• Kubernetes rightsizing from VPA recommendations: Reads Vertical Pod Autoscaler (VPA) recommendations and updates resource requests when the improvement is significant.
• Idle and waste detection: Identifies zero-CPU workloads and unmounted PVCs to reclaim compute and storage, while gating high-risk deletions behind human approval.
• Database maintenance scheduling: Runs ANALYZE for stale table statistics and schedules compaction/OPTIMIZE for Iceberg tables with many small files.
• Optimization auditing and reporting: Produces an audit trail and a weekly report with category-level savings attribution.

Quick Start

Use the aiops-platform-optimization-agent skill to run an optimization cycle that discovers platform improvements, applies low-risk actions, requests approval for high-risk changes, and outputs a weekly optimization report.