Update on the cloud waste scanner I posted about last month - added AI/ML rules across AWS, Azure, and GCP Shared the hygiene rule list here about a month ago. Wanted to post an update since the scope has changed quite a bit. **What's new since then:** Added AI/ML rules across all three providers, opt-in with `--category ai`. These target resources that look quiet from a billing dashboard but are still running and accruing charges. AWS (6 new rules, 19 total): * SageMaker inference endpoints — InService with no invocations * SageMaker notebook instances — InService but no recent activity * SageMaker Studio apps (JupyterServer, KernelGateway) — InService but idle * SageMaker training jobs running well past a normal threshold * Bedrock Provisioned Throughputs — InService with no request traffic * EC2 GPU instances with near-zero utilization Azure (5 new rules, 17 total): * AML compute clusters with a baseline node floor and no observed job activity for 14+ days * AML compute instances in Running state with no…