HTML to Image Lambda Function

This project contains an AWS Lambda function that renders HTML content to an image using Puppeteer and the @sparticuz/chromium package. The project is deployed using AWS CDK.

Features

• Render HTML content to PNG or JPEG images
• Return images as base64-encoded strings or upload to S3 and return a signed URL
• Configure viewport size, image quality, and other rendering options
• Deployed with API Gateway for easy access

Development Tools

This project was developed using:

• Cursor.sh v0.47.6 - An AI-powered code editor
• Amazon Q CLI - AWS's AI-powered developer tool
• AWS CDK - Infrastructure as Code tool for AWS

Architecture

The project includes:

1. HTML to Image Lambda Function: Renders HTML to images using Puppeteer and Chromium
2. S3 Bucket: Stores generated images when requested
3. API Gateway: Provides HTTP endpoints to access the Lambda functions
4. Simple HTML Renderer: A basic Lambda function that returns HTML content (for testing)

API Usage

HTML to Image Endpoint

Endpoint: POST /render-image

Request Body:

{
  "html": "base64-encoded-html-content",
  "outputType": "base64", // or "s3"
  "imageFormat": "png", // or "jpeg"
  "imageQuality": 80, // For JPEG only (1-100)
  "viewportWidth": 1280,
  "viewportHeight": 800,
  "fullPage": false // Set to true to capture the full scrollable page
}

Response (base64 mode):

{
  "message": "Screenshot generated successfully",
  "imageFormat": "png",
  "image": "base64-encoded-image"
}

Response (S3 mode):

{
  "message": "Screenshot saved to S3",
  "url": "signed-s3-url",
  "key": "s3-object-key"
}

HTML Renderer Endpoint

Endpoint: GET /render?name=YourName

Returns a simple HTML page with the provided name parameter.

Deployment

To deploy this project:

npm install
npm run build
npx cdk deploy

Local Development

To test the function locally:

npm run build
npx cdk synth

Notes

• The S3 bucket is configured to automatically delete objects after 7 days
• The Lambda function has a 30-second timeout and 2GB of memory
• For production use, consider adjusting the S3 bucket's removal policy and lifecycle rules

MCP Server

This includes an MCP server, which we can run as:

.venv/bin/mcp run server.py