Babak d’Abbaschian

Babak d’Abbaschian

7ASecurity

ponti

Elie Steinbock ~ getinboxzero.com

Jay ⛽️

Jay ⛽️

@jelanifuel

15 Jun 2025

But first why should you learn this? 1. Because you can use claude 4 to build them with ease. If you need something done in n8n but can't figure it out or there isn't a node for it, just build a lambda function. 2. Takes n8n/relevance etc to next level because your agents can invoke these python functions. With Python you have the ability to do SO much more but it's really not available in n8n. 3. Most importantly, Lambda functions are serverless meaning you only pay for them when they are running which is often for milliseconds so the cost is insanely cheap vs having a server running all the time for one thing. Most API's are using Lambda functions and charging you a premium. But you can build your own and cut cost. This particular guide will share how you can create AWS images using Lambda and then Auto store them in a public s3 bucket for you to share/download. But the point is you can create all types of functions using Claude 4 once you see how easy it is. Alright lets go. Complete AWS Lambda Setup Guide - OpenAI Image Generator From Brand New AWS Account to Working Function PART 1: INITIAL AWS ACCOUNT SETUP Step 1: Access AWS Console Go to aws.amazon.com/ Click "Sign in to the Console" Enter your credentials and sign in Make sure you're in the correct region (top-right corner)Recommended: us-east-1 (N. Virginia) or us-west-2 (Oregon) PART 2: CREATE S3 BUCKET Step 2: Create S3 Bucket for Images In AWS Console search bar, type "S3" Click on "S3" service Click "Create bucket" (big orange button) Configure Basic Settings: Bucket name: openai-images-bucket-[your-name-or-number]Example: openai-images-bucket-john123 Must be globally unique, so add something personal AWS Region: Choose same region as you selected earlier Leave other settings as default for now Configure Public Access:Scroll down to "Block Public Access settings for this bucket" UNCHECK "Block all public access" ❌ Check the acknowledgment box that says "I acknowledge that the current settings..." Scroll to bottom and click "Create bucket" Step 3: Configure Bucket for Public Image Access Click on your newly created bucket name (not the checkbox, the actual name) Click the "Permissions" tab Scroll down to "Bucket policy" Click "Edit" Copy and paste this policy (replace YOUR-BUCKET-NAME with your actual bucket name): json{ "Version": "2012-10-17", "Statement": [ { "Sid": "PublicReadGetObject", "Effect": "Allow", "Principal": "*", "Action": "s3:GetObject", "Resource": "arn:aws:s3:::YOUR-BUCKET-NAME/*" } ] } Click "Save changes" ✅ CHECKPOINT: Your S3 bucket is now ready to store public images. PART 3: CREATE IAM ROLE Step 4: Create IAM Role for Lambda In AWS Console search bar, type "IAM" Click on "IAM" service In left sidebar, click "Roles" Click "Create role" (blue button) Select Trusted Entity:Select "AWS service" (should be selected by default) Use case: Select "Lambda" Click "Next" Add Permissions - First Policy:In search box, type: AWSLambdaBasicExecutionRole Check the checkbox next to it Click "Next" Name and Create Role:Role name: OpenAI-Image-Lambda-Role Description: Role for OpenAI image generation Lambda function Click "Create role" Step 5: Add S3 Permissions to Role Click on the role name OpenAI-Image-Lambda-Role you just created Click "Add permissions" dropdown → "Create inline policy" Click "JSON" tab Delete the existing JSON and paste this (replace YOUR-BUCKET-NAME): json{ "Version": "2012-10-17", "Statement": [ { "Effect": "Allow", "Action": [ "s3:PutObject", "s3:PutObjectAcl" ], "Resource": "arn:aws:s3:::YOUR-BUCKET-NAME/*" } ] } Click "Next" Policy name: S3-Upload-Policy Click "Create policy" ✅ CHECKPOINT: Your IAM role now has permissions to run Lambda and upload to S3. PART 4: CREATE LAMBDA FUNCTION Step 6: Create Lambda Function In AWS Console search bar, type "Lambda" Click on "Lambda" service Click "Create function" (orange button) Configure Function:Select "Author from scratch" (should be selected) Function name: openai-image-generator Runtime: Select "Python 3.11" from dropdown Architecture: Leave as x86_64 Change Execution Role:Expand "Change default execution role" Select "Use an existing role" Existing role: Select OpenAI-Image-Lambda-Role from dropdown Click "Create function" Step 7: Configure Lambda Settings Wait for function to be created (you'll see a green success banner) Click on "Configuration" tab In left sidebar, click "General configuration" Click "Edit" Update settings:Timeout: Change to 1 min 30 sec (or type 90 seconds) Memory: Change to 512 MB Click "Save" Step 8: Add Environment Variables Still in Configuration tab, click "Environment variables" in left sidebar Click "Edit" Click "Add environment variable" First Variable: Key: OPENAI_API_KEY Value: your-actual-openai-api-key-here Get this from platform.openai.com/api-keys Click "Add environment variable" again Second Variable: Key: S3_BUCKET_NAME Value: your-actual-bucket-name-here (the name you created in Step 2) Click "Save" ✅ CHECKPOINT: Your Lambda function is configured with proper timeout, memory, and environment variables. PART 5: DEPLOY THE CODE Step 9: Add the Python Code Click on "Code" tab You'll see lambda_function.py in the file explorer on the left Click on lambda_function.py Select ALL the default code (Ctrl A or Cmd A) Delete it Copy and paste your working code (the code from your document) === ACTUAL CODE === import json import boto3 import urllib.request import urllib.parse import urllib.error import base64 import uuid import os from typing import Dict, Any, List def lambda_handler(event: Dict[str, Any], context: Any) -> Dict[str, Any]: """ AWS Lambda function to generate images using OpenAI's gpt-image-1 API and upload them to S3. Expected event structure: { "prompt": "A description of the image to generate", "size": "1024x1024" (optional, defaults to "auto"), "n": 1 (optional, number of images to generate), "background": "auto" (optional: "auto", "transparent", "opaque"), "quality": "auto" (optional: "auto", "high", "medium", "low"), "output_format": "png" (optional: "png", "jpeg", "webp"), "output_compression": 100 (optional, 0-100), "moderation": "auto" (optional: "auto", "low") } """ try: # Parse input parameters body = event.get('body') if isinstance(body, str): body = json.loads(body) elif body is None: body = event prompt = body.get('prompt') if not prompt: return { 'statusCode': 400, 'body': json.dumps({'error': 'prompt is required'}) } # Extract optional parameters with defaults size = body.get('size', 'auto') n = body.get('n', 1) background = body.get('background', 'auto') quality = body.get('quality', 'auto') output_format = body.get('output_format', 'png') output_compression = body.get('output_compression', 100) moderation = body.get('moderation', 'auto') # Validate parameters if n < 1 or n > 10: return { 'statusCode': 400, 'body': json.dumps({'error': 'n must be between 1 and 10'}) } # Environment variables openai_api_key = os.environ.get('OPENAI_API_KEY') s3_bucket = os.environ.get('S3_BUCKET_NAME') if not openai_api_key: return { 'statusCode': 500, 'body': json.dumps({'error': 'OPENAI_API_KEY environment variable not set'}) } if not s3_bucket: return { 'statusCode': 500, 'body': json.dumps({'error': 'S3_BUCKET_NAME environment variable not set'}) } # Generate images using OpenAI API image_urls = generate_images( api_key=openai_api_key, prompt=prompt, size=size, n=n, background=background, quality=quality, output_format=output_format, output_compression=output_compression, moderation=moderation ) # Upload images to S3 and get URLs s3_urls = upload_images_to_s3(image_urls, s3_bucket, output_format) return { 'statusCode': 200, 'body': json.dumps({ 'success': True, 'images': s3_urls, 'count': len(s3_urls) }) } except Exception as e: print(f"Error: {str(e)}") return { 'statusCode': 500, 'body': json.dumps({'error': str(e)}) } def generate_images(api_key: str, prompt: str, size: str, n: int, background: str, quality: str, output_format: str, output_compression: int, moderation: str) -> List[str]: """ Generate images using OpenAI's gpt-image-1 API. Returns: List of base64-encoded image strings """ url = "api.openai.com/v1/images/gen…" payload = { "model": "gpt-image-1", "prompt": prompt, "n": n, "size": size, "background": background, "quality": quality, "output_format": output_format, "output_compression": output_compression, "moderation": moderation } # Remove None values payload = {k: v for k, v in payload.items() if v is not None} # Convert payload to JSON data = json.dumps(payload).encode('utf-8') # Create request req = urllib.request.Request( url, data=data, headers={ "Authorization": f"Bearer {api_key}", "Content-Type": "application/json" } ) try: with urllib.request.urlopen(req, timeout=60) as response: if response.status != 200: error_data = response.read().decode('utf-8') raise Exception(f"OpenAI API error: {response.status} - {error_data}") result = json.loads(response.read().decode('utf-8')) except urllib.error.HTTPError as e: error_data = e.read().decode('utf-8') raise Exception(f"OpenAI API error: {e.code} - {error_data}") except urllib.error.URLError as e: raise Exception(f"Network error: {e.reason}") # Extract base64 images from response images = [] for image_data in result.get('data', []): if 'b64_json' in image_data: images.append(image_data['b64_json']) else: raise Exception("Expected base64 image data not found in response") return images def upload_images_to_s3(base64_images: List[str], bucket_name: str, output_format: str) -> List[str]: """ Upload base64-encoded images to S3 and return public URLs. Args: base64_images: List of base64-encoded image strings bucket_name: S3 bucket name output_format: Image format (png, jpeg, webp) Returns: List of S3 URLs """ s3_client = boto3.client('s3') s3_urls = [] # Determine file extension and content type if output_format == 'jpeg': extension = 'jpg' content_type = 'image/jpeg' elif output_format == 'webp': extension = 'webp' content_type = 'image/webp' else: # png extension = 'png' content_type = 'image/png' for i, base64_image in enumerate(base64_images): try: # Decode base64 image image_data = base64.b64decode(base64_image) # Generate unique filename filename = f"generated-images/{uuid.uuid4()}.{extension}" # Upload to S3 s3_client.put_object( Bucket=bucket_name, Key=filename, Body=image_data, ContentType=content_type, CacheControl='max-age=31536000' # Cache for 1 year ) # Generate public URL s3_url = f"https://{bucket_name}.s3.amazonaws.com/{filename}" s3_urls.append(s3_url) except Exception as e: print(f"Error uploading image {i}: {str(e)}") raise Exception(f"Failed to upload image {i} to S3: {str(e)}") return s3_urls # Optional: Helper function for testing locally def test_locally(): """ Test function for local development. Set environment variables before running. """ test_event = { "prompt": "A serene mountain landscape with a lake at sunset", "size": "1024x1024", "n": 1, "quality": "high", "output_format": "png" } result = lambda_handler(test_event, None) print(json.dumps(result, indent=2)) if __name__ == "__main__": test_locally() === END OF CODE === Click "Deploy" (orange button) Wait for "Successfully updated the function" message ✅ CHECKPOINT: Your code is now deployed to Lambda. PART 6: TEST THE FUNCTION Step 10: Create and Run Test Still in Code tab, click "Test" (next to Deploy button) Configure test event:Select "Create new event" Event name: basic-test Template: Leave as "hello-world" Replace the JSON with this test case: json{ "prompt": "A serene mountain landscape with a crystal clear lake reflecting the sky", "size": "1024x1024", "n": 1 } Click "Save" Click "Test" (this will run your function) Step 11: Check Test Results If Successful, you'll see: Green banner saying "Execution result: succeeded" Response should look like: json{ "statusCode": 200, "body": { "success": true, "images": [ "your-bucket-name.s3.amazonaw…" ], "count": 1 } } Click on the S3 URL to verify the image was generated correctly If you get errors, see troubleshooting section below. ✅ CHECKPOINT: Your function successfully generates images and uploads them to S3! PART 7: OPTIONAL - CREATE API GATEWAY Step 12: Create API Gateway (Optional - for HTTP access) In AWS Console search bar, type "API Gateway" Click on "API Gateway" service Under "REST API", click "Build" (not the Private one) Configure API:Choose the protocol: REST Create new API: Selected API name: OpenAI-Image-API Description: API for generating images with OpenAI Endpoint Type: Regional Click "Create API" Step 13: Create API Resource and Method In API Gateway console, click "Actions" dropdown Select "Create Resource" Configure Resource:Resource Name: generate-image Resource Path: /generate-image (auto-filled) Enable API Gateway CORS: ✅ Check this Click "Create Resource" With /generate-image selected, click "Actions" → "Create Method" Select "POST" from dropdown → Click checkmark Setup Method:Integration type: Lambda Function Use Lambda Proxy integration: ✅ Check this Lambda Region: Your region Lambda Function: Start typing openai and select your function Click "Save" Click "OK" on the permission popup Step 14: Deploy API Click "Actions" → "Deploy API" Deployment stage: Select "[New Stage]" Stage name: prod Click "Deploy" Copy the Invoke URL (you'll need this for HTTP requests) ✅ CHECKPOINT: You now have an HTTP API endpoint for your image generator! PART 8: VERIFICATION AND TESTING Step 15: Test Different Scenarios Test 1 - Basic Test (in Lambda console): json{ "prompt": "A red apple on a white background" } Test 2 - Multiple Images: json{ "prompt": "A futuristic city skyline", "size": "1536x1024", "n": 3, "quality": "high" } Test 3 - API Gateway Test (using curl or Postman): bashcurl -X POST your-api-id.execute-api.regi… \ -H "Content-Type: application/json" \ -d '{ "prompt": "A magical forest with glowing mushrooms", "size": "1024x1024", "n": 1 }' Step 16: Verify S3 Images Go to S3 console Click on your bucket You should see a generated-images/ folder Click on it to see your generated images Click on any image file to view/download it TROUBLESHOOTING COMMON ISSUES Error: "OPENAI_API_KEY environment variable not set" Solution: Go back to Step 8 and double-check your environment variables. Error: "S3_BUCKET_NAME environment variable not set" Solution: Verify the bucket name in environment variables matches exactly. Error: "Access Denied" when uploading to S3 Solution: Check IAM role has the S3 policy (Step 5) Verify bucket name in IAM policy matches your actual bucket name Error: OpenAI API errors Solution: Verify your OpenAI API key is valid Check you have credits/quota on your OpenAI account Ensure your API key has access to gpt-image-1 model Error: "Task timed out after X seconds" Solution: Increase Lambda timeout (Step 7) Try with smaller images or fewer images (n=1) Images not accessible via URL Solution: Check bucket policy is set correctly (Step 3) Verify bucket has public read access enabled FINAL CHECKLIST S3 bucket created with public read access IAM role created with Lambda and S3 permissions Lambda function created with correct runtime and role Environment variables set (OPENAI_API_KEY, S3_BUCKET_NAME) Lambda timeout set to 90 seconds Lambda memory set to 512 MB Code deployed successfully Function tested with sample prompt Images generated and accessible via S3 URLs API Gateway created (optional) API tested via HTTP request (if using API Gateway) 🎉 CONGRATULATIONS! Your OpenAI image generation Lambda function is now fully operational! WHAT YOU'VE BUILT You now have: ✅ Lambda function that generates images using OpenAI's gpt-image-1 API ✅ S3 storage for generated images with public access ✅ Proper security with IAM roles and policies ✅ HTTP API (optional) for external access ✅ Error handling and logging via CloudWatch Your function accepts prompts and returns public S3 URLs to generated images! Now to use your new function in n8n you can search for the Lambda node in n8n and then connect your AWS connect which will allow you to select the functions you created. Then just select invoke as the operation and then send the data the function needs to run in the json input area. But if you get stuck Claude 4 can help you solve this as well. For example for this function and step by step guide I used this simple prompt: "i want to create a python lambda function that calls the openai gpt-image-1 api to generate an image. Here are the details of the api call: We only want to call the gpt-image-1. The prompt and size will be sent to you. Once you get back the base64 from the openai call we need to convert the base64/upload it to a s3 bucket so we can view the image and return the s3 link(s)." Claude did the rest. I just clicked a few buttons and had a running function. Anyway, that's it. Have fun and take your building to the next level.

724

Cache Control: An Easy P4 Vulnerability

Smart CDN | Supabase Docs