Cloud Cost Optimization Analyzer

Authority checks:

• AWS: Cost Explorer API enabled, ce:GetCostAndUsage permission
• Azure: Cost Management API access, Reader role on subscriptions
• GCP: Cloud Billing API enabled, billing.accounts.getSpendingInformation permission

Source citations (accessed 2025-10-25T22:42:30-04:00):

• AWS Cost Management: https://aws.amazon.com/aws-cost-management/
• Azure Cost Optimization: https://azure.microsoft.com/solutions/cost-optimization/
• GCP Cost Management: https://cloud.google.com/cost-management
• FinOps Framework: https://www.finops.org/framework/
• FinOps Principles: https://www.finops.org/framework/principles/

Procedure

Tier 1 (≤2k tokens): Quick Cost Health Check

Goal: Identify top 3 cost optimization opportunities in <5 minutes.

Steps:

1. Fetch cost summary for specified time_range

   • Group by service/resource type
   • Calculate total spend and trend (% change from previous period)

2. Quick waste scan (top offenders only)

   • Stopped instances still attached to storage
   • Unattached EBS volumes / Azure managed disks / GCP persistent disks
   • Unused Elastic IPs / Public IPs / Static external IPs
   • Load balancers with zero traffic (last 7 days)

3. Commitment coverage check

   • Calculate % of compute spend covered by reserved instances / savings plans / committed use discounts
   • If <60% coverage → flag as optimization opportunity

4. Output quick wins (3 highest impact items)

   • Example: "Delete 15 unattached EBS volumes → save $450/month"
   • Example: "Purchase EC2 Savings Plan (3-year) → save $2,400/month"
   • Example: "Rightsize 8 over-provisioned RDS instances → save $1,200/month"

Token budget checkpoint: ~1.5k tokens for API calls, analysis, and output formatting.

Tier 2 (≤6k tokens): Comprehensive Cost Analysis

Goal: Generate detailed, actionable cost optimization plan with quantified savings.

Extends T1 with:

5. Rightsizing analysis

   • Fetch CloudWatch / Azure Monitor / GCP Monitoring metrics (CPU, memory, network utilization)
   • Identify resources with <20% utilization over 90th percentile
   • Recommend downsizing (example: m5.2xlarge → m5.xlarge saves $150/month)
   • Calculate savings per resource: (current_price - recommended_price) * hours_per_month

6. Reserved instance / Savings plan optimization

   • Analyze historical usage patterns (30d minimum, 90d preferred)
   • Identify stable workloads eligible for commitments
   • Calculate break-even point: upfront_cost / monthly_savings
   • Recommend commitment term (1-year vs 3-year) based on usage stability
   • AWS: Compare Compute Savings Plans vs EC2 RIs vs convertible RIs
   • Azure: Compare Reserved VM Instances vs Azure Hybrid Benefit
   • GCP: Compare Committed Use Discounts (CUD) vs Sustained Use Discounts (SUD)

7. Storage optimization

   • Identify candidates for lifecycle policies (hot → cool → archive)
   • AWS S3: Recommend Intelligent-Tiering or S3 Glacier transitions
   • Azure Blob: Recommend tiering to Cool or Archive
   • GCP Cloud Storage: Recommend Nearline or Coldline classes
   • Calculate savings: (current_storage_cost - optimized_cost) * TB_stored

8. Network cost optimization

   • Identify cross-region / cross-AZ traffic (expensive)
   • Recommend VPC endpoints / Private Link to avoid NAT gateway costs
   • Flag public internet egress (most expensive path)

9. Cost anomaly detection

   • Calculate baseline spend (average + 2 std deviations)
   • Flag spikes >20% above baseline
   • Attribute anomaly to specific service/resource

10. FinOps maturity assessment (basic)

    • Tag compliance: % resources with required cost allocation tags
    • Budget variance: actual vs budgeted spend
    • Commitment utilization: % of purchased RI/SP actually used
    • Waste ratio: idle_cost / total_cost

Authority sources (accessed 2025-10-25T22:42:30-04:00):

• AWS Reserved Instances: https://aws.amazon.com/ec2/pricing/reserved-instances/
• AWS Savings Plans: https://aws.amazon.com/savingsplans/ (up to 72% savings)
• Azure Reserved Instances: https://learn.microsoft.com/azure/cost-management-billing/reservations/
• Azure Hybrid Benefit: https://azure.microsoft.com/pricing/hybrid-benefit/ (up to 36% for Windows Server)
• GCP Committed Use Discounts: https://cloud.google.com/compute/docs/instances/committed-use-discounts-overview
• FinOps "Crawl, Walk, Run" maturity: https://www.finops.org/framework/

Output: JSON report with sections: cost_summary, quick_wins (T1), rightsizing_recommendations, commitment_recommendations, storage_optimization, network_optimization, anomalies, finops_maturity_score.

Token budget checkpoint: ~5k tokens (includes T1 + extended analysis + detailed output).

Tier 3 (not implemented for T2 skill)

Reserved for future enhancements: predictive cost forecasting, ML-based anomaly detection, multi-account/org-wide consolidation, custom FinOps policies.

Decision Rules

When to abort:

• Cost data source returns 403/401 → insufficient permissions; emit setup instructions
• Cost data empty or <7 days → insufficient data for analysis
• All optimization targets already at maximum efficiency (rare) → report "no action needed"

Ambiguity thresholds:

• Rightsizing confidence: Only recommend if utilization <20% for 90% of time period (avoids false positives from bursty workloads)
• Commitment recommendations: Require 90d minimum history AND <10% variance in daily usage to recommend 3-year term
• Anomaly detection: Only flag if >20% deviation AND >$100 absolute difference (avoid noise)

Prioritization logic:

1. Highest ROI first: savings_per_month / implementation_effort
   • Effort scale: Low (delete unused) < Medium (rightsize) < High (migrate architecture)
2. Quick wins: Zero-downtime changes (stop unused, delete orphaned) rank highest
3. Risk-adjusted: Downsizing production workloads requires manual approval; rank lower

FinOps principle application (accessed 2025-10-25T22:42:30-04:00):

Per FinOps Foundation principles (https://www.finops.org/framework/principles/):

• "Everyone takes ownership": Tag all recommendations with owning team/project via cost allocation tags
• "Centralized optimization": Reserved instance / savings plan purchases centralized; this skill generates recommendations for central FinOps team
• "Variable cost model as opportunity": Emphasize autoscaling and spot instances as cost-saving strategies

Output Contract

Schema (JSON):

{
  "cost_analysis_report": {
    "period": "2025-09-25 to 2025-10-25",
    "cloud_provider": "aws",
    "total_spend": 47500.32,
    "waste_identified": 14250.10,
    "savings_potential": {
      "monthly": 12800.00,
      "annual": 153600.00,
      "percentage": 26.9
    }
  },
  "quick_wins": [
    {
      "category": "unused_resources",
      "description": "Delete 12 unattached EBS volumes",
      "monthly_savings": 360.00,
      "implementation_effort": "low",
      "risk_level": "none"
    }
  ],
  "rightsizing_recommendations": [
    {
      "resource_id": "i-0a1b2c3d4e5f6g7h8",
      "resource_type": "ec2_instance",
      "current_type": "m5.2xlarge",
      "recommended_type": "m5.xlarge",
      "utilization_avg": 18.5,
      "monthly_savings": 150.00,
      "confidence": "high"
    }
  ],
  "commitment_recommendations": [
    {
      "service": "ec2_compute",
      "plan_type": "compute_savings_plan",
      "term": "3_year",
      "upfront": "partial",
      "monthly_commitment": 5000.00,
      "monthly_savings": 1200.00,
      "break_even_months": 4.2
    }
  ],
  "waste_inventory": [
    {
      "resource_id": "vol-0123456789abcdef",
      "type": "ebs_volume_unattached",
      "idle_days": 45,
      "monthly_cost": 30.00
    }
  ],
  "action_plan": [
    {
      "priority": 1,
      "action": "Delete unused resources",
      "impact": "high",
      "effort": "low",
      "items_count": 27,
      "monthly_savings": 810.00
    }
  ]
}

Required fields: cost_analysis_report (with total_spend, savings_potential), action_plan (prioritized).

Optional fields: rightsizing_recommendations, commitment_recommendations (only if applicable).

Examples

# Example: AWS cost optimization for over-provisioned workload