AWS CDK TypeScript
Overview
Use this skill to build AWS infrastructure in TypeScript with reusable constructs, safe defaults, and a validation-first delivery loop.
When to Use
Use this skill when:
- Creating or refactoring a CDK app, stack, or reusable construct in TypeScript
- Choosing between L1, L2, and L3 constructs
- Building serverless, networking, or security-focused AWS infrastructure
- Wiring multi-stack applications and environment-aware deployments
- Validating infrastructure changes with
cdk synth, tests,cdk diff, andcdk deploy
Instructions
1. Project Initialization
# Create a new CDK app
npx cdk init app --language typescript
# Project structure
my-cdk-app/
├── bin/
│ └── my-cdk-app.ts # App entry point (instantiates stacks)
├── lib/
│ └── my-cdk-app-stack.ts # Stack definition
├── test/
│ └── my-cdk-app.test.ts # Tests
├── cdk.json # CDK configuration
├── tsconfig.json
└── package.json
2. Core Architecture
import { App, Stack, StackProps, CfnOutput, RemovalPolicy } from 'aws-cdk-lib';
import { Construct } from 'constructs';
import * as s3 from 'aws-cdk-lib/aws-s3';
// Define a reusable stack
class StorageStack extends Stack {
public readonly bucketArn: string;
constructor(scope: Construct, id: string, props?: StackProps) {
super(scope, id, props);
const bucket = new s3.Bucket(this, 'DataBucket', {
versioned: true,
encryption: s3.BucketEncryption.S3_MANAGED,
removalPolicy: RemovalPolicy.RETAIN,
blockPublicAccess: s3.BlockPublicAccess.BLOCK_ALL,
});
this.bucketArn = bucket.bucketArn;
new CfnOutput(this, 'BucketName', { value: bucket.bucketName });
}
}
// App entry point
const app = new App();
new StorageStack(app, 'DevStorage', {
env: { account: process.env.CDK_DEFAULT_ACCOUNT, region: 'us-east-1' },
tags: { Environment: 'dev' },
});
new StorageStack(app, 'ProdStorage', {
env: { account: '123456789012', region: 'eu-west-1' },
tags: { Environment: 'prod' },
terminationProtection: true,
});
app.synth();
3. Construct Levels
| Level | Description | Use When |
|-------|-------------|----------|
| L1 (Cfn*) | Direct CloudFormation mapping, full control | Need properties not exposed by L2 |
| L2 | Curated with sensible defaults and helper methods | Standard resource provisioning (recommended) |
| L3 (Patterns) | Multi-resource architectures | Common patterns like LambdaRestApi |
// L1 — Raw CloudFormation
new s3.CfnBucket(this, 'L1Bucket', { bucketName: 'my-l1-bucket' });
// L2 — Sensible defaults + grant helpers
const bucket = new s3.Bucket(this, 'L2Bucket', { versioned: true });
bucket.grantRead(myLambda);
// L3 — Multi-resource pattern
new apigateway.LambdaRestApi(this, 'Api', { handler: myLambda });
4. CDK Lifecycle Commands
cdk synth # Synthesize CloudFormation template
cdk diff # Compare deployed vs local changes
cdk deploy # Deploy stack(s) to AWS
cdk deploy --all # Deploy all stacks
cdk destroy # Tear down stack(s)
cdk ls # List all stacks in the app
cdk doctor # Check environment setup
5. Recommended Delivery Loop
-
Model the stack
- Start with L2 constructs and extract repeated logic into custom constructs.
-
Run
cdk synth- Checkpoint: synthesis succeeds with no missing imports, invalid props, missing context, or unresolved references.
- If it fails: fix the construct configuration or context values, then rerun
cdk synth.
-
Run infrastructure tests
- Checkpoint: assertions cover IAM scope, stateful resources, and critical outputs.
- If tests fail: update the stack or test expectations, then rerun the test suite.
-
Run
cdk diff- Checkpoint: review IAM broadening, resource replacement, export changes, and deletes on stateful resources.
- If the diff is risky: adjust names, dependencies, or
RemovalPolicy, then reruncdk diff.
-
Run
cdk deploy- Checkpoint: the stack reaches
CREATE_COMPLETEorUPDATE_COMPLETE. - If deploy fails: inspect CloudFormation events, fix quotas, permissions, export conflicts, or bootstrap issues, then retry
cdk deploy.
- Checkpoint: the stack reaches
-
Verify runtime outcomes
- Confirm stack outputs, endpoints, alarms, and integrations behave as expected before moving on.
6. Cross-Stack References
// Stack A exports a value
class NetworkStack extends Stack {
public readonly vpc: ec2.Vpc;
constructor(scope: Construct, id: string, props?: StackProps) {
super(scope, id, props);
this.vpc = new ec2.Vpc(this, 'Vpc', { maxAzs: 2 });
}
}
// Stack B imports it via props
interface AppStackProps extends StackProps {
vpc: ec2.Vpc;
}
class AppStack extends Stack {
constructor(scope: Construct, id: string, props: AppStackProps) {
super(scope, id, props);
new lambda.Function(this, 'Fn', {
runtime: lambda.Runtime.NODEJS_20_X,
handler: 'index.handler',
code: lambda.Code.fromAsset('lambda'),
vpc: props.vpc,
});
}
}
// Wire them together
const network = new NetworkStack(app, 'Network');
new AppStack(app, 'App', { vpc: network.vpc });
Examples
Example 1: Serverless API
import * as cdk from 'aws-cdk-lib';
import * as lambda from 'aws-cdk-lib/aws-lambda';
import * as apigateway from 'aws-cdk-lib/aws-apigateway';
import * as dynamodb from 'aws-cdk-lib/aws-dynamodb';
class ServerlessApiStack extends cdk.Stack {
constructor(scope: Construct, id: string, props?: cdk.StackProps) {
super(scope, id, props);
const table = new dynamodb.Table(this, 'Items', {
partitionKey: { name: 'id', type: dynamodb.AttributeType.STRING },
billingMode: dynamodb.BillingMode.PAY_PER_REQUEST,
removalPolicy: cdk.RemovalPolicy.DESTROY,
});
const fn = new lambda.Function(this, 'Handler', {
runtime: lambda.Runtime.NODEJS_20_X,
handler: 'index.handler',
code: lambda.Code.fromAsset('lambda'),
environment: { TABLE_NAME: table.tableName },
});
table.grantReadWriteData(fn);
new apigateway.LambdaRestApi(this, 'Api', { handler: fn });
}
}
Example 2: CDK Assertion Test
import { Template } from 'aws-cdk-lib/assertions';
import { App } from 'aws-cdk-lib';
import { ServerlessApiStack } from '../lib/serverless-api-stack';
test('creates DynamoDB table with PAY_PER_REQUEST', () => {
const app = new App();
const stack = new ServerlessApiStack(app, 'TestStack');
const template = Template.fromStack(stack);
template.hasResourceProperties('AWS::DynamoDB::Table', {
BillingMode: 'PAY_PER_REQUEST',
});
template.resourceCountIs('AWS::Lambda::Function', 1);
});
Best Practices
- One concern per stack — Separate network, compute, storage, and monitoring.
- Prefer L2 constructs — Drop to
Cfn*only when you need unsupported properties. - Set explicit environments — Pass
envwith account and region; avoid implicit production targets. - Use grant helpers — Prefer
.grant*()over handwritten IAM where possible. - Review the diff before deploy — Treat IAM expansion, replacement, and deletes as mandatory checkpoints.
- Test infrastructure — Cover critical resources with fine-grained assertions.
- Avoid hardcoded values — Use context, parameters, or environment variables.
- Use the right
RemovalPolicy—RETAINfor production data,DESTROYonly for disposable environments.
Constraints and Warnings
- CloudFormation limits — Max 500 resources per stack; split large apps into multiple stacks
- Synthesis is not deployment —
cdk synthonly generates templates;cdk deployapplies changes - Cross-stack references create CloudFormation exports; removing them requires careful ordering
- Stateful resources (RDS, DynamoDB, S3 with data) — Always set
removalPolicy: RETAINin production - Bootstrap required — Run
cdk bootstraponce per account/region before first deploy - Asset bundling — Lambda code and Docker images are uploaded to the CDK bootstrap bucket
References
Detailed implementation guides are available in the references/ directory:
- Core Concepts — App lifecycle, stacks, constructs, environments, assets
- Serverless Patterns — Lambda, API Gateway, DynamoDB, S3 events, Step Functions
- Networking & VPC — VPC design, subnets, NAT, security groups, VPC endpoints
- Security Hardening — IAM, KMS, Secrets Manager, WAF, compliance
- Testing Strategies — Assertions, snapshots, integration tests, CDK Nag