I built a CLI tool to solve a specific pain point: Identifying expensive GPU instances that are technically "running" but doing zero actual work (low CPU/Network).

Features:

Agentless (uses CloudWatch/Azure Monitor metrics) Multi-cloud (AWS, Azure, GCP) Outputs a "Remediation Manifest" (Terraform state rm commands) It’s designed to be run in a CI/CD pipeline as a daily "Garbage Collector" for dev environments.

I'd love feedback on the detection logic (currently threshold-based <5% CPU).