Cost Bloodhound
Find the $15K/month in orphaned resources nobody knows about
Hunts through your cloud accounts for wasted spend — orphaned EBS volumes, idle load balancers, oversized instances, forgotten snapshots, unused Elastic IPs. Generates a cleanup report with estimated savings and the commands to fix each one.
INGREDIENTS
PROMPT
Create a skill called "Cost Bloodhound". Scan my cloud accounts for wasted spend. Using the AWS CLI (and/or GCP/Azure equivalents): 1. Find orphaned EBS volumes: `aws ec2 describe-volumes --filters Name=status,Values=available` 2. Find unused Elastic IPs: `aws ec2 describe-addresses` cross-referenced with running instances 3. Find idle load balancers: ALBs/NLBs with zero healthy targets or negligible traffic 4. Find old snapshots: snapshots older than 90 days not referenced by any AMI 5. Find oversized instances: CloudWatch CPU avg < 10% over 14 days (and memory too, if memory metrics are available) 6. Find idle RDS instances: connections < 1 over 7 days 7. Find unattached EBS volumes across ALL regions For each finding, show: - Resource ID, type, region, and tags - Monthly cost estimate - Recommended action (delete, downsize, or investigate) - The exact CLI command to remediate (with dry-run where available) Generate a summary with total estimated monthly savings.
How It Works
Uses your cloud CLI to scan for the common patterns that waste 28-32% of
the average cloud bill. No third-party tool needed — just AWS/GCP/Azure
CLI access.
What You Get
- Orphaned resources: unattached EBS volumes, unused Elastic IPs, idle load balancers
- Oversized instances: comparing CPU usage, and memory usage where metrics exist, to instance type
- Old snapshots and AMIs that are no longer referenced
- Idle RDS instances and over-provisioned databases
- Estimated monthly savings for each finding
- Cleanup commands (with dry-run options) for each resource
- Cost trend analysis by service and tag
Setup Steps
- Ensure your Claw has read access to AWS/GCP/Azure CLI
- Tell it which accounts and regions to scan
- Review the waste report and estimated savings
- Run cleanup commands (or Terraform import to bring resources under management)
Tips
- Start with the biggest wins: orphaned EBS volumes and idle load balancers are usually the top offenders
- Check across ALL regions — forgotten resources in ap-southeast-1 are a classic
- Run monthly to catch drift before the bill surprises you
- Tag findings so you can track cleanup progress
- The savings estimates are conservative — actual savings are usually higher