cloud-aws-architect_skill

Overview

This skill designs multi-service AWS solutions across compute, storage, networking, and serverless with cost optimization, security hardening, and alignment to the AWS Well-Architected Framework. It produces quick architecture recommendations (T1) and production-ready designs with infrastructure-as-code (T2) including CloudFormation or CDK artifacts, cost estimates, and migration plans.

How this skill works

Validate workload inputs (requirements, workload_type, deployment_tier, regions, budget_constraints) and apply decision thresholds to select compute, storage, networking, and integration services. For T1 it returns a compact architecture, justification, rough monthly cost estimate, and Well-Architected alignment. For T2 it produces a detailed design, IaC templates or CDK constructs, security hardening steps, monitoring, and a 1/3-year TCO with optimization recommendations.

When to use it

• Designing an application or system that spans compute, storage, and networking on AWS
• Performing a Well-Architected review or optimization for cost, security, or reliability
• Planning multi-region deployments or disaster recovery strategies
• Migrating on-premises or other-cloud workloads to AWS with IaC
• Choosing between serverless, containers, or EC2 for a new workload

Best practices

• Classify the workload (web-app, data-processing, real-time, batch, machine-learning, hybrid) before selecting services
• Prefer least-privilege IAM roles, KMS-managed encryption, and account-level S3 Block Public Access
• Use Savings Plans, reserved capacity, and spot instances where appropriate to reduce costs
• Design multi-AZ by default; choose multi-region only when required for RTO/RPO
• Automate deployments with CloudFormation or CDK and enforce tagging for cost allocation

Example use cases

• Rapid T1 recommendation for a web application: API Gateway + Lambda or ALB + ECS Fargate with S3 static assets
• T2 production design for a data platform: S3 data lake, Glue/EMR, Redshift, VPC with Transit Gateway and monitoring
• Migration plan for on-prem web + database: discovery → rehost/refactor → DMS for DB migration → DNS cutover
• Cost optimization audit: identify rightsizing, Savings Plans, S3 lifecycle rules, and spot/auto-scaling opportunities