cost-optimization-audit

# Cost Optimization Audit ## Step 1: Gather context Ask the user: > What workload or AWS environment would you like me to audit for cost optimization? Please share: > - **Architecture overview** (services used, rough monthly spend if known) > - **Traffic patterns** (steady, spiky, predictable growth, seasonal) > - **Commitment status** (Savings Plans, Reserved Instances, existing contracts) > - **Budget constraints or targets** (optional) If context is already provided, proceed directly. ## Step 2: Identify waste Classify each finding by severity: - 🔴 **High** — significant ongoing waste (>20% of category spend or >$500/month) - 🟡 **Medium** — moderate waste (5-20% of category spend or $100-500/month) - 🟢 **Low** — minor waste or optimization opportunity Look for: - **Idle resources** — unattached EBS volumes, unused Elastic IPs, idle load balancers, stopped instances with attached storage - **Over-provisioned resources** — instances with <20% CPU/memory utilization, over-sized RDS instances, over-provisioned DynamoDB capacity - **Orphaned resources** — old snapshots, unused AMIs, stale log groups, abandoned S3 buckets - **Redundant data transfer** — cross-AZ traffic that could be avoided, NAT Gateway costs for S3/DynamoDB (use VPC endpoints) ## Step 3: Evaluate pricing models Assess: - Are Savings Plans or Reserved Instances used for steady-state workloads? - Could Spot Instances cover fault-tolerant or batch workloads? - Would serverless (Lambda, Fargate, Aurora...