cost-optimization

name	cost-optimization
description	Cloud cost optimization with Infracost for Terraform, ccusage for Claude Code tokens, and AWS Cost Explorer patterns
triggers	["cost optimization","infracost","ccusage","cloud costs","aws costs","budget","spending"]

Cost Optimization

Track and optimize costs across infrastructure (AWS/Terraform) and AI tools (Claude Code).

Claude Code Usage Tracking

# View usage analytics from local JSONL logs
npx ccusage

# Daily breakdown
npx ccusage --daily

# Monthly summary
npx ccusage --monthly

# Live monitoring of current session
npx ccusage --live

Infracost (Terraform Cost Estimation)

# Show cost breakdown for current Terraform
infracost breakdown --path terraform/

# Compare cost of changes (in CI)
infracost diff --path terraform/ --compare-to infracost-base.json

# Generate baseline for comparison
infracost breakdown --path terraform/ --format json --out-file infracost-base.json

CI Integration

- name: Infracost
  uses: infracost/actions/setup@v3
- run: infracost diff --path terraform/ --format json --compare-to /tmp/base.json --out-file /tmp/diff.json
- uses: infracost/actions/comment@v3
  with:
    path: /tmp/diff.json
    behavior: update

AWS Cost Patterns

Right-sizing Checklist

• EC2: Check CPU <20% avg -> downsize instance type
• RDS: Check connections <50% -> consider smaller instance or Aurora Serverless
• EKS: Check pod resource requests vs actual usage -> adjust requests/limits
• S3: Enable Intelligent-Tiering for unpredictable access patterns
• Lambda: Optimize memory allocation (use AWS Lambda Power Tuning)

Budget Alerts

resource "aws_budgets_budget" "monthly" {
  name         = "jarvis-monthly"
  budget_type  = "COST"
  limit_amount = "500"
  limit_unit   = "USD"
  time_unit    = "MONTHLY"

  notification {
    comparison_operator = "GREATER_THAN"
    threshold           = 80
    threshold_type      = "PERCENTAGE"
    notification_type   = "ACTUAL"
    subscriber_email_addresses = ["<[e-mail]>"]
  }
}

Cost Optimization Quick Wins

1. Use Spot instances for non-critical workloads (60-90% savings)
2. Reserved Instances or Savings Plans for steady-state (30-40% savings)
3. S3 lifecycle policies to transition to Glacier after 90 days
4. Delete unattached EBS volumes and unused Elastic IPs
5. Use opusplan in Claude Code (Opus for planning, Sonnet for coding)