Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Compare AWS, GCP, and Azure to select the best cloud provider based on workload requirements, cost, compliance, and migration complexity.
.claude/skills/williamzujkowski-cloud-provider-selection-advisor/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | 418% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 165% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 257% | 0% |
| case-22 | ✗→✓ | ▲ Improved | 195% | 0% |
| case-11 | ✓→✗ | ▼ Worse | 132% | 0% |
Trigger conditions:
Not for:
Time normalization:
NOW_ET using NIST/time.gov semantics (America/New_York, ISO-8601): 2025-10-26T18:00:00-04:00NOW_ET for all citation access datesInput validation:
requirements must include at least one priority: cost, performance, security, complianceworkload_type must be one of: web-app, data-processing, real-time, batch, machine-learning, hybridpriorities must be ranked list (highest priority first)constraints if provided must specify hard requirements (e.g., "must support eu-west-1", "must have FedRAMP High")Source freshness:
Fast path for 80% of cloud selection decisions:
| Category | AWS | GCP | Azure | |----------|-----|-----|-------| | Compute (VMs) | EC2 | Compute Engine | Virtual Machines | | Compute (Serverless) | Lambda | Cloud Functions | Functions | | Compute (Containers) | ECS/Fargate | Cloud Run | Container Apps | | Compute (K8s) | EKS | GKE | AKS | | Object Storage | S3 | Cloud Storage | Blob Storage | | Block Storage | EBS | Persistent Disk | Managed Disks | | Relational DB | RDS/Aurora | Cloud SQL | SQL Database | | NoSQL | DynamoDB | Firestore/Bigtable | Cosmos DB | | Data Warehouse | Redshift | BigQuery | Synapse Analytics | | Networking | VPC | VPC | Virtual Network | | Load Balancer | ALB/NLB | Cloud Load Balancing | Load Balancer/App Gateway | | CDN | CloudFront | Cloud CDN | Front Door |
AWS:
GCP:
Azure:
Choose AWS if:
Choose GCP if:
Choose Azure if:
From on-premises:
From AWS to GCP/Azure:
From GCP to AWS/Azure:
Abort conditions:
For complex cloud selection with detailed comparison:
Compute tier:
Storage tier:
Database tier:
Compute cost example (4 vCPU, 16GB RAM, 730 hrs/month):
Storage cost example (1TB object storage, 100K requests/month):
Data transfer costs (100GB outbound/month):
FedRAMP:
HIPAA:
Data residency:
Developer experience:
CI/CD integration:
Monitoring and observability:
When to use multi-cloud:
Multi-cloud patterns:
Multi-cloud challenges:
High lock-in risk:
Low lock-in risk:
Mitigation strategies:
Abort conditions:
Note: This skill implements T1 (quick recommendations) and T2 (detailed comparison with multi-cloud strategy) tiers only. T2 provides comprehensive cloud provider comparison with service-by-service analysis, cost breakdowns, migration planning, and multi-cloud architecture patterns. For highly specialized scenarios requiring deeper vendor consultation (custom pricing negotiations, enterprise agreements, dedicated support contracts), engage directly with AWS, GCP, or Azure sales teams.
Future T3 considerations:
Primary cloud selection:
AWS if:
GCP if:
Azure if:
Multi-cloud if:
Cost priority:
Performance priority:
Ease of migration:
Ambiguity handling:
Stop conditions:
Required fields (all tiers):
json{ "recommendation": { "primary_provider": "aws | gcp | azure | multi-cloud", "rationale": ["array of 3-5 reasons for recommendation"], "confidence": "high | medium | low" }, "service_mapping": { "compute": { "aws": "service name", "gcp": "service name", "azure": "service name", "recommended": "provider::service with justification" }, "storage": "...", "database": "...", "networking": "..." }, "cost_comparison": { "aws_monthly_usd": "number", "gcp_monthly_usd": "number", "azure_monthly_usd": "number", "cheapest": "provider", "optimization_strategies": ["array of discount/commitment options"] }, "considerations": { "vendor_lock_in": "low | medium | high with services at risk", "migration_complexity": "low | medium | high (if applicable)", "compliance": ["array of compliance frameworks met"], "regional_coverage": "assessment of region availability" } }
Additional T2 fields:
json{ "detailed_service_comparison": [ { "category": "compute | storage | database | networking", "aws": {"service": "name", "pros": [], "cons": []}, "gcp": {"service": "name", "pros": [], "cons": []}, "azure": {"service": "name", "pros": [], "cons": []} } ], "cost_breakdown": { "compute": {"aws": "number", "gcp": "number", "azure": "number"}, "storage": {"aws": "number", "gcp": "number", "azure": "number"}, "database": {"aws": "number", "gcp": "number", "azure": "number"}, "networking": {"aws": "number", "gcp": "number", "azure": "number"} }, "migration_plan": { "source": "current infrastructure", "target": "recommended provider", "complexity": "low | medium | high", "timeline": "string (e.g., 3-6 months)", "phases": ["array of migration phases"], "tools": ["array of migration tools"] }, "multi_cloud_architecture": { "pattern": "best-of-breed | primary-dr | data-distribution | geographic", "providers": ["array of providers used"], "service_distribution": {"provider": ["array of services"]} } }
yaml# T1 Example: Startup Web Application Requirements: workload_type: web-app priorities: [cost, developer-experience, scalability] constraints: {region: us-east, compliance: none} Recommendation: GCP Rationale: - Lowest cost with sustained use discounts (automatic 20-30% savings) - Developer-friendly gcloud CLI and Cloud Shell - Cloud Run for serverless containers (better than Lambda for web apps) - Free tier generous (Cloud Run 2M requests/month) Service Mapping: Compute: Cloud Run (serverless containers, auto-scale 0-1000) Database: Cloud SQL PostgreSQL (managed, automated backups) Storage: Cloud Storage (static assets, 11 9s durability) CDN: Cloud CDN (integrated with Cloud Run) Cost Estimate: GCP: $45/month (Cloud Run $15, Cloud SQL $25, Storage $5) AWS: $63/month (Lambda $20, RDS $35, S3 $8) Azure: $73/month (Functions $18, SQL Database $45, Blob $10) Migration: N/A (greenfield project)
Token budgets (enforced):
Accuracy requirements:
Determinism:
Official Cloud Documentation (accessed 2025-10-26T18:00:00-04:00):
Pricing Calculators:
Well-Architected Frameworks:
Migration Tools:
Third-Party Comparisons:
Related Skills:
cloud-aws-architect, cloud-gcp-architect, cloud-azure-architectfinops-cost-analyzercloud-multicloud-advisorcloud-kubernetes-integrator| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 19,752 | 9,564 | -52% | 1 | 1 | 0% | 4,639 | 8,008 | +73% | 0 | 0 | — |
case-02 | pass→pass | 27,651 | 25,850 | -7% | 1 | 1 | 0% | 6,338 | 11,748 | +85% | 0 | 0 | — |
case-03 | pass→pass | 9,417 | 5,551 | -41% | 1 | 1 | 0% | 2,288 | 7,310 | +219% | 0 | 0 | — |
case-04 | pass→pass | 10,972 | 10,385 | -5% | 1 | 1 | 0% | 2,483 | 8,262 | +233% | 0 | 0 | — |
case-05 | pass→pass | 15,033 | 9,024 | -40% | 1 | 1 | 0% | 3,346 | 8,057 | +141% | 0 | 0 | — |
case-06 | pass→pass | 29,793 | 13,852 | -54% | 1 | 1 | 0% | 5,995 | 8,643 | +44% | 0 | 0 | — |
case-07 | pass→pass | 19,293 | 12,176 | -37% | 1 | 1 | 0% | 3,974 | 8,543 | +115% | 0 | 0 | — |
case-08 | fail→pass | 7,104 | 10,948 | +54% | 1 | 1 | 0% | 1,624 | 8,418 | +418% | 0 | 0 | — |
case-09 | pass→pass | 18,476 | 14,951 | -19% | 1 | 1 | 0% | 3,952 | 9,311 | +136% | 0 | 0 | — |
case-10 | pass→pass | 13,539 | 12,048 | -11% | 1 | 1 | 0% | 2,724 | 8,654 | +218% | 0 | 0 | — |
case-11 | pass→fail | 19,226 | 13,663 | -29% | 1 | 1 | 0% | 3,845 | 8,938 | +132% | 0 | 0 | — |
case-12 | pass→pass | 14,530 | 8,516 | -41% | 1 | 1 | 0% | 2,895 | 7,760 | +168% | 0 | 0 | — |
case-13 | pass→pass | 16,962 | 14,985 | -12% | 1 | 1 | 0% | 3,706 | 9,233 | +149% | 0 | 0 | — |
case-20 | pass→pass | 18,202 | 14,060 | -23% | 1 | 1 | 0% | 3,721 | 9,128 | +145% | 0 | 0 | — |
case-14 | fail→pass | 14,088 | 7,578 | -46% | 1 | 1 | 0% | 2,829 | 7,486 | +165% | 0 | 0 | — |
case-15 | fail→fail | 15,251 | 15,392 | +1% | 1 | 1 | 0% | 3,063 | 9,211 | +201% | 0 | 0 | — |
case-16 | fail→pass | 13,323 | 17,143 | +29% | 1 | 1 | 0% | 2,505 | 8,942 | +257% | 0 | 0 | — |
case-17 | pass→pass | 11,775 | 9,459 | -20% | 1 | 1 | 0% | 2,378 | 7,752 | +226% | 0 | 0 | — |
case-18 | fail→fail | 23,791 | 15,289 | -36% | 1 | 1 | 0% | 4,198 | 8,880 | +112% | 0 | 0 | — |
case-19 | pass→pass | 18,439 | 17,141 | -7% | 1 | 1 | 0% | 4,310 | 9,875 | +129% | 0 | 0 | — |
case-21 | pass→pass | 14,423 | 17,681 | +23% | 1 | 1 | 0% | 3,576 | 10,157 | +184% | 0 | 0 | — |
case-22 | fail→pass | 16,890 | 12,285 | -27% | 1 | 1 | 0% | 2,879 | 8,481 | +195% | 0 | 0 | — |
case-23 | pass→pass | 14,761 | 16,042 | +9% | 1 | 1 | 0% | 3,090 | 9,215 | +198% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +13 percentage points is the difference between those two pass rates over the 23 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.