Guide · Agent skill
How to use the AWS Solution Architect agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The aws-solution-architect skill is a packaged set of instructions and Python scripts that lets an AI agent act as a startup-focused AWS architect. It gathers your application requirements, suggests one of four standard AWS designs, drafts infrastructure-as-code such as CloudFormation or CDK snippets, and reviews spending for cost-saving ideas. This guide is for readers who have never used agent skills, and by the end you will know how to install the skill, trigger it with a plain request, and deploy safely.
AWS Solution Architectalirezarezvani/claude-skills · MIT Download the skill Quick start
- 1Install the skill. Download the ZIP from this page and unzip it into ~/.claude/skills/ or a project's .claude/skills/ folder, or upload it under Settings → Capabilities → Skills.
- 2Check prerequisites. Confirm Python 3 is installed. Install and configure the AWS CLI with an account's credentials only if you plan to deploy.
- 3Describe your application. Ask in plain language, including users, budget, team experience and compliance needs.
- 4Review the recommended pattern. Check that the suggested pattern and service stack match your team's skills and requirements.
- 5Read the generated template. Open the CloudFormation YAML and confirm each resource before you deploy anything.
- 6Validate and deploy. Run validate-template, then create the stack with CAPABILITY_IAM if the template creates IAM roles.
- 7Review costs. Provide your current resource inventory and monthly spend to get right-sizing and savings recommendations.
What an agent skill is and what this one does
An agent skill is a folder containing a SKILL.md file of instructions, plus optional reference documents, sample files and scripts that the agent can use when a request matches the skill's description.
This skill is built for startups. Its workflow has six steps: gather requirements, design the architecture, generate IaC templates, review costs, deploy, and validate and handle failures. It covers Lambda, API Gateway, DynamoDB, ECS, Aurora and cost optimization.
- Gathers app type, expected users, budget, team experience, compliance needs and availability requirements
- Recommends one of four patterns: serverless web, event-driven microservices, three-tier, or GraphQL backend
- Drafts a serverless CloudFormation template or CDK/Terraform snippets
- Reviews a current AWS setup for right-sizing, Savings Plans and storage tier ideas
- Provides deployment and troubleshooting CLI commands
Install the aws-solution-architect skill
Download the ZIP for the skill from this page and unzip it. The skill folder contains SKILL.md, the scripts folder, the references folder and the assets folder.
For a personal setup, unzip the skill into ~/.claude/skills/. To limit it to one project, unzip it into the project's .claude/skills/ folder instead.
Alternatively, upload the ZIP in the Claude apps under Settings → Capabilities → Skills.
- Personal: ~/.claude/skills/
- Single project: .claude/skills/ inside your project folder
- Claude apps: Settings → Capabilities → Skills, then upload the ZIP
Check what you need before you start
The bundled scripts need Python 3. You need the AWS CLI and an AWS account with credentials only for the deploy and validation steps. CDK or Terraform are optional and only needed if you choose those deployment routes.
The scripts run on your machine and the skill uses the network, so review what it will run before you approve it.
- Python 3 to run the bundled scripts
- AWS CLI and AWS credentials, needed only for deploy and validation
- Optional: CDK or Terraform
- Network use is listed for this skill
Trigger the skill with natural requests
You do not need to type a command to use the skill. Describe what you want in plain language that matches its purpose: designing serverless architecture, creating CloudFormation templates, optimizing AWS costs, setting up CI/CD pipelines, or migrating to AWS.
Give the agent concrete numbers where you can, such as user count, monthly budget and compliance needs, because the design step uses them.
- "Design a serverless MVP backend for a mobile app with 1000 users."
- "Design a scalable architecture for a SaaS platform with 50k users."
- "Optimize my AWS setup to reduce costs by 30%. Current spend: $3000/month."
Review the design and generated templates
The architecture designer script produces a recommended pattern, service stack, cost estimate and pros and cons. Confirm the pattern fits your team's operational maturity and compliance requirements before generating templates.
The serverless stack script creates a CloudFormation YAML file for API Gateway with Lambda, a DynamoDB table, a Cognito user pool, IAM roles and CloudWatch logging. The reference file architecture_patterns.md contains full pattern specifications.
- Pattern options: Serverless Web, Event-Driven Microservices, Three-Tier, GraphQL Backend
- Templates are generic starting points and may miss required controls
- Cost figures are estimates
python scripts/serverless_stack.py --app-name my-app --region us-east-1Deploy and validate safely
Before deploying, validate the template. Deployment runs AWS CLI commands that create IAM roles and billable resources, so read the templates first. The stack deploy command uses CAPABILITY_IAM because the template creates IAM roles.
If stack creation fails, check the failure reason in the stack events, review CloudWatch Logs, fix the template, delete the failed stack, and redeploy.
- Validate first with aws cloudformation validate-template
- Check IAM permission errors by verifying --capabilities CAPABILITY_IAM and role trust policies
- Resource limit errors: request a quota increase through the Service Quotas console
aws cloudformation validate-template --template-body file://template.yamlWhat people use it for
- Designing a serverless MVP backend with Lambda, API Gateway, DynamoDB and Cognito for a small mobile app
- Planning a scalable SaaS setup with ECS Fargate, Aurora Serverless and CloudFront
- Getting a first-pass review of AWS spend before commissioning a deeper audit
- Generating a starter CloudFormation template to adapt for a new project
Frequently asked questions
How do I install the aws-solution-architect skill?
Download the ZIP from this page and unzip it into ~/.claude/skills/ for personal use, or into a project's .claude/skills/ folder. You can also upload the ZIP in the Claude apps under Settings → Capabilities → Skills.
What do I need to run the AWS Solution Architect skill?
You need Python 3 to run the bundled scripts. The AWS CLI and an AWS account with credentials are needed only for deploy and validation steps, and CDK or Terraform are optional.
Is the aws-solution-architect skill free?
The skill is licensed under MIT, as listed on this page. Any AWS resources you deploy are billed by AWS to your account.
Can I use this skill for HIPAA, SOC 2 or GDPR workloads?
Not without expert review. The skill accepts compliance requirements as an input, but its templates are generic starting points and may miss required controls.
Will the skill deploy to my AWS account automatically?
The workflow includes AWS CLI deploy commands that create IAM roles and billable resources, so you run them yourself after reading the templates. Read the generated templates before deploying to a live account.
Are the cost estimates accurate?
The cost figures are estimates, and the scripts use simple built-in heuristics. Do not rely on them as a verified cost forecast.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you are a startup engineer or small team that needs a sensible first AWS setup, a starter serverless API template, or a quick first-pass review of AWS spend. Pick something else, or bring in an expert, for production or regulated workloads, or if you need a verified cost forecast.
Download the skill