aws-solution-architect
by alirezarezvaniaws-solution-architect is a Claude skill for Cloud Architecture planning on AWS, helping design startup-ready serverless, ECS, Aurora, DynamoDB, and event-driven stacks with cost estimates, service tradeoffs, IaC direction, and phased implementation guidance.
This skill scores 80/100, meaning it is a solid listing candidate for directory users who want an agent to produce AWS startup architecture recommendations with less guesswork than a generic prompt. It has a clear use case, structured workflow, scripts, and practical references, though adopters should expect to do their own setup and production validation.
- Clear trigger description covers serverless architecture design, CloudFormation/IaC generation, cost optimization, CI/CD, and AWS migration scenarios.
- Operational workflow asks for concrete requirements such as traffic, budget, team experience, compliance, availability, and RPO/RTO before design.
- Includes supporting scripts, sample input/output, and reference guides for architecture patterns, best practices, and service selection.
- No install command or README is provided, so users must infer setup from the skill path and script examples.
- The AWS recommendations appear heuristic and startup-oriented; users should validate generated architectures, costs, and compliance choices before production use.
Overview of aws-solution-architect skill
What aws-solution-architect is for
aws-solution-architect is a Claude skill for designing startup-friendly AWS architectures with service recommendations, cost estimates, and infrastructure-as-code direction. It is best suited for Cloud Architecture work where the goal is not just “pick AWS services,” but produce a defensible architecture pattern, implementation phases, and cost-aware tradeoffs.
Best-fit users and decisions
This aws-solution-architect skill fits founders, startup engineers, platform teams, and consultants who need a first-pass AWS design for SaaS platforms, mobile backends, web apps, microservices, data pipelines, or migrations. It is especially useful when you must compare Lambda/API Gateway/DynamoDB serverless patterns against ECS Fargate, Aurora, ElastiCache, EventBridge, Step Functions, or multi-region options.
What makes it different from a generic prompt
The repository includes practical decision material: references/architecture_patterns.md, references/service_selection.md, references/best_practices.md, plus Python scripts for architecture recommendation, serverless stack generation, and cost optimization. That gives the agent a clearer workflow than a broad “design an AWS architecture” prompt: gather requirements, select a pattern, estimate costs, generate IaC-style output, then review risks.
Where it is strongest and weakest
It is strongest for early architecture planning, cost-sensitive startup designs, serverless-first systems, and structured AWS service selection. It should not be treated as a final security review, compliance attestation, or production deployment plan without human AWS review. Pricing, regional availability, quotas, and compliance controls should be verified against current AWS documentation before implementation.
How to Use aws-solution-architect skill
aws-solution-architect install and repository path
Install the skill from the GitHub repository with:
npx skills add alirezarezvani/claude-skills --skill aws-solution-architect
The skill lives at:
engineering-team/skills/aws-solution-architect
After installing, read SKILL.md first for the workflow. Then inspect:
assets/sample_input.jsonfor the expected requirement shapeassets/expected_output.jsonfor output structurereferences/architecture_patterns.mdfor pattern selectionreferences/service_selection.mdfor service tradeoffsreferences/best_practices.mdfor reliability, security, and cost guidancescripts/architecture_designer.py,scripts/cost_optimizer.py, andscripts/serverless_stack.pyfor implementation logic
Inputs the skill needs before it can help
For strong aws-solution-architect usage, provide more than an app idea. Include application type, expected users, requests per second, region, monthly budget, team size, AWS experience, data size, compliance needs, availability target, and special features such as authentication, real-time updates, multi-region, or autoscaling.
A weak prompt is:
Design AWS architecture for my SaaS.
A stronger prompt is:
Use aws-solution-architect for Cloud Architecture. Design an AWS architecture for a B2B SaaS in
us-east-1with 50,000 expected users, 100 RPS average, 500 GB relational data, GDPR requirements, $1,500/month target budget, 5 engineers, intermediate AWS experience, authentication, high availability, autoscaling, no real-time features, and no multi-region requirement. Compare serverless and ECS/Aurora options, estimate monthly cost, list risks, and propose phased implementation.
This works better because the skill’s decision logic depends on traffic, budget, team capability, compliance, and workload shape.
Suggested workflow for real projects
Start by asking for requirement gaps before asking for a final design. Then request a pattern recommendation, service stack, cost breakdown, tradeoffs, and IaC direction. For example:
- “Ask me missing AWS architecture requirements before designing.”
- “Recommend the best pattern using
references/architecture_patterns.md.” - “Explain why the rejected patterns are weaker for this case.”
- “Generate a phased implementation plan with CloudFormation, CDK, or Terraform assumptions.”
- “Review cost, security, scaling, and operational risks.”
If you already have AWS resources, include a current inventory and spend so the cost optimizer path is relevant.
Practical tips for better output
Be explicit about constraints that change service choice: latency target, long-running jobs, database consistency, compliance scope, disaster recovery, expected traffic spikes, deployment preference, and whether the team can operate containers or databases. If you need IaC, specify the format: CloudFormation YAML, AWS CDK TypeScript, or Terraform HCL. If you care about startup cost control, ask for “MVP now, scale later” rather than only a final-state architecture.
aws-solution-architect skill FAQ
Is aws-solution-architect suitable for beginners?
Yes, if the user can provide basic product and traffic assumptions. The skill includes service selection references that explain when to use Lambda, ECS Fargate, Aurora, DynamoDB, API Gateway, SQS, EventBridge, Cognito, CloudFront, and related AWS services. Beginners should still ask it to explain tradeoffs and implementation phases rather than only generating templates.
When should I not use this skill?
Do not use it as the only source for regulated production systems, complex enterprise landing zones, deep networking, Kubernetes-heavy platforms, or final compliance decisions. It can surface HIPAA, GDPR, SOC 2, availability, and security considerations, but it does not replace a certified AWS architect, threat model, Well-Architected Review, or current AWS pricing validation.
How is it different from asking Claude directly?
A direct prompt can produce a plausible AWS diagram, but the aws-solution-architect skill adds a repeatable workflow, example input/output schema, architecture pattern references, service decision matrices, and scripts that encode recommendation and cost-optimization behavior. That makes outputs more consistent and easier to audit.
Does it generate deployable infrastructure?
It can guide IaC generation and the expected output references CloudFormation, AWS CDK, and Terraform-style deliverables. Treat generated infrastructure as a starting point. Before deployment, review IAM scope, VPC design, secrets handling, encryption, logging, alarms, backup policy, environment separation, and region-specific pricing.
How to Improve aws-solution-architect skill
Improve aws-solution-architect results with clearer constraints
The fastest way to improve aws-solution-architect output is to provide measurable constraints. Replace “scalable and cheap” with “support 100 RPS now, 1,000 RPS in 12 months, stay under $1,500/month, tolerate 4-hour RTO, use managed services where possible.” The skill can then choose between serverless, ECS Fargate, Aurora, DynamoDB, and event-driven patterns with less guesswork.
Ask for tradeoffs, not only recommendations
For architecture work, the rejected options matter. Ask the skill to explain why it did not choose EKS, EC2, Aurora, DynamoDB, App Runner, or Lambda for a given workload. This exposes hidden assumptions around cold starts, 15-minute Lambda limits, operational complexity, database access patterns, predictable traffic, and team experience.
Iterate from first draft to implementation plan
After the first architecture, request a second pass focused on risks: cost hotspots, single points of failure, IAM risks, observability gaps, data migration, rollback strategy, and quota limits. Then ask for phased delivery: foundation, core services, security and monitoring, CI/CD, and scale-out. This mirrors the structure in assets/expected_output.json and produces a more actionable plan.
Validate scripts and references before trusting output
Before using the skill for a serious design, inspect the scripts and references in the repository. architecture_designer.py shows what requirement fields affect recommendations. cost_optimizer.py indicates what cost categories it considers. service_selection.md and best_practices.md reveal the assumptions behind the guidance. This makes the aws-solution-architect guide easier to adapt to your AWS account, region, compliance level, and deployment toolchain.