aws
π―SkillInstallation
npx skills add g1joshi/agent-skills --skill awsRun it in a terminal; it asks which AI coding tools to install it for (Claude Code, Codex, Cursor and more).
Uses skills.sh Β· Node.js 22.20+
After installing
Ask for related work and the agent uses it on its own, or call it yourself.
/awsInstall by hand
Copy the whole skill folder from GitHub to the place below. Copying only SKILL.md can leave out files the skill uses.
Open the skill folder β- This project
.claude/skills/aws/ - All projects
~/.claude/skills/aws/
These paths are for Claude Code. For other tools, the command above puts it in the right place.
SKILL.md
View on GitHub βExpert Amazon Web Services (AWS) cloud assistance covering IAM, EC2, S3, ECS/EKS, Lambda, VPC, and CloudWatch. Use when designing, deploying, and managing scalable AWS cloud architectures.
AWS
Amazon Web Services (AWS) is a comprehensive cloud platform providing scalable compute, serverless architectures, managed databases, and enterprise AI infrastructure.
When to Use
- Enterprise Cloud Infrastructure: Global compute (EC2, ECS, EKS), storage (S3), and database (Aurora, DynamoDB) solutions.
- Serverless & Event-Driven Applications: AWS Lambda, EventBridge, SQS, SNS, and API Gateway for zero-idle cost scaling.
- High-Security Multi-Account Architectures: AWS Organizations, IAM Identity Center, Control Tower, and KMS.
- Edge Computing & Global Content Delivery: CloudFront CDN, Route 53 DNS, and AWS WAF.
Quick Start
# Verify credentials and caller identity
aws sts get-caller-identity
# List S3 buckets
aws s3 ls
# Sync local assets directory to private S3 bucket
aws s3 sync ./dist s3://my-app-assets-bucket/ --delete
Core Concepts
Serverless Architecture with AWS Lambda & SQS
Event-driven microservice processing incoming messages:
import json
import boto3
import os
dynamodb = boto3.resource('dynamodb')
table = dynamodb.Table(os.environ['ORDERS_TABLE'])
def handler(event, context):
processed_count = 0
for record in event['Records']:
body = json.loads(record['body'])
order_id = body['order_id']
amount = body['amount']
# Persist transaction to DynamoDB
table.put_item(Item={
'PK': f"ORDER#{order_id}",
'amount': str(amount),
'status': 'CONFIRMED',
'timestamp': record['attributes']['ApproximateFirstReceiveTimestamp']
})
processed_count += 1
return {
'statusCode': 200,
'body': json.dumps({'processed': processed_count})
}
Least-Privilege IAM Policy
Scoping permissions strictly to resource ARNs:
{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "AllowS3InvoiceBucketAccess",
"Effect": "Allow",
"Action": ["s3:GetObject", "s3:PutObject"],
"Resource": "arn:aws:s3:::corporate-invoices-prod/*"
},
{
"Sid": "AllowKMSDecryption",
"Effect": "Allow",
"Action": "kms:Decrypt",
"Resource": "arn:aws:kms:us-east-1:123456789012:key/abc-123",
"Condition": {
"StringEquals": {
"kms:CallerAccount": "123456789012"
}
}
}
]
}
High-Availability VPC Network Architecture
Subnet layout for zero-trust network segregation:
[Internet Gateway]
β
[Public Subnets (NAT Gateway, ALB)] βββΊ 10.0.1.0/24, 10.0.2.0/24
β
[Private Application Subnets (ECS, EKS, Lambda)] βββΊ 10.0.10.0/24, 10.0.11.0/24
β
[Isolated Database Subnets (Aurora, Redis)] βββΊ 10.0.20.0/24, 10.0.21.0/24
Common Patterns
Least-Privilege IAM Policy for Application Service Roles
Problem: Granting wildcard * permissions exposes AWS accounts to severe compromise.
Solution: Scoped IAM policy restricting actions to specific resources:
{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": ["s3:GetObject", "s3:PutObject"],
"Resource": "arn:aws:s3:::my-production-bucket/uploads/*"
},
{
"Effect": "Allow",
"Action": "dynamodb:GetItem",
"Resource": "arn:aws:dynamodb:us-east-1:123456789012:table/UserSessions"
}
]
}
Best Practices
Do:
- Organize workloads into multi-account architectures using AWS Organizations and AWS Control Tower.
- Enforce least-privilege IAM policies without wildcard actions (
"Action": "*") or wildcard resources. - Enable S3 Block Public Access and Default KMS Encryption across all storage buckets.
- Deploy databases and application containers into private subnets with egress via NAT Gateways.
Don't:
- Use AWS root account credentials for daily management or API tasks; lock with hardware MFA.
- Leave CloudWatch log groups with indefinite retention; configure explicit retention policies (e.g. 30 days).
- Hardcode AWS credentials; use IAM Roles for EC2/ECS/Lambda or OIDC for GitHub Actions.
Troubleshooting
| Error | Cause | Solution |
|---|---|---|
AccessDenied (Service: Amazon S3; Status Code: 403) | IAM user/role lacks permission or S3 Bucket Policy blocks access. | Verify IAM policy permissions and check bucket Block Public Access settings. |
ExpiredToken: The security token included in the request is expired | Temporary STS credentials (AWS SSO or assumed role) have expired. | Re-authenticate: aws sso login or refresh STS tokens. |
VPC Resource Timeout / No internet access in Lambda | Lambda placed in private subnet without NAT Gateway. | Route private subnet 0.0.0.0/0 traffic through a NAT Gateway in a public subnet. |
References
More from this repository10
A template skill providing a starter framework for creating new agent skills with structured sections for triggers, quick start examples, and capability descriptions.
Skill
Skill
Skill
Skill
Skill
Skill
Skill
Skill
Skill