MemeOn

Foundational project for the MemeOn.ai marketing site and serverless API.

Project layout

• web/ – Vite-powered frontend that displays Hello MemeOn.ai and calls the serverless API.
• api/ – TypeScript Lambda handler plus an Express dev server that serves /api routes on port 3001, returning a JSON description (from api/src/api-description.ts) for any unimplemented path.
• infra/terraform/ – Terraform configuration for S3, CloudFront, ACM, API Gateway, Lambda and Route53.
• .github/workflows/ – CI/CD automation for build and deploy (see below).

Prerequisites

• Node.js 20+
• npm 10+
• Terraform 1.8+
• AWS CLI v2 with credentials that can manage S3, CloudFront, Lambda, API Gateway, ACM and Route53.

Local development

Install dependencies once:

npm install

Then launch both the Vite dev server and the API server locally:

npm run dev

The site runs on http://localhost:5173 and proxies /api requests to http://localhost:3001.

Building

Build the frontend and package the Lambda bundle:

npm run build

The Lambda package is written to api/lambda.zip and is referenced by Terraform.

Infrastructure

Terraform is organised under infra/terraform.

1. Generate the Lambda bundle: npm run bundle:api.
2. Ensure the S3 bucket memeon.ai exists. You can create it with the AWS CLI:

   aws s3api create-bucket --bucket memeon.ai --region us-west-2 --create-bucket-configuration LocationConstraint=us-west-2
   

3. Initialise and plan:

   cd infra/terraform
   terraform init
   terraform plan -var='route53_zone_id=Z123456789ABC'
   

4. Apply when ready:

   terraform apply -var='route53_zone_id=Z123456789ABC'
   

The configuration provisions:

• S3 bucket for the static site with CORS set to mirror the Masky project.
• CloudFront distribution with behaviours for the site and /api/* requests.
• CloudFront function to redirect www.memeon.ai to memeon.ai.
• ACM certificate in us-east-1 for memeon.ai and www.memeon.ai.
• API Gateway HTTP endpoint integrated with the memeon-api Lambda function.
• Route53 records for apex and www.

Certificate validation requires DNS. Terraform creates the CNAME records automatically in the hosted zone you supply.

Deployment workflow

A GitHub Actions workflow (.github/workflows/deploy.yml) builds the frontend, uploads assets to S3, invalidates the CloudFront cache, and updates the Lambda function code using AWS CLI commands. Static deployment identifiers (including the AWS region) live in config/deploy.json. Configure the following repository secrets before running the workflow:

• AWS_ACCESS_KEY_ID
• AWS_SECRET_ACCESS_KEY
• AWS_REGION (optional if awsRegion is present in config/deploy.json)

Trigger the workflow with pushes to main or manually from the Actions tab.