AI Content Auto-Poster

Automated social media content generator that creates and posts AI-generated images with captions to Facebook using AWS Lambda and Hugging Face APIs.

🌟 Features

• AI-Generated Content: Uses Hugging Face models for text and image generation
• Automated Posting: Scheduled daily posts via AWS EventBridge
• Serverless Architecture: Runs entirely on AWS Lambda (no servers to manage)
• Infrastructure as Code: Complete Terraform configuration for easy deployment

🏗️ Architecture

EventBridge (Cron) → Lambda Function → Hugging Face API → Facebook Graph API

• AWS Lambda: Serverless function execution
• AWS EventBridge: Scheduled triggers (daily at 2 PM UTC)
• Hugging Face API:
  • Text generation (GLM-4.7-Flash)
  • Image generation (Stable Diffusion XL)
• Facebook Graph API: Posts content to your Facebook page

📋 Prerequisites

• AWS Account with CLI configured
• Hugging Face account and API token
• Facebook Page and API credentials
• Terraform installed
• Python 3.13
• PowerShell (for Windows) or Bash (for Linux/Mac)

🚀 Setup

1. Clone the Repository

git clone <your-repo-url>
cd ai-content-auto-poster

2. Get API Credentials

Hugging Face Token

1. Go to Hugging Face Settings
2. Create a new token with read permissions
3. Copy the token

Facebook Page Token

1. Go to Facebook Developers
2. Create an app and add Facebook Login product
3. Get a Page Access Token for your Facebook Page
4. Get your Page ID from your Facebook Page settings

3. Configure Terraform Variables

Create a terraform.tfvars file:

lambda_function_name = "ai-content-auto-poster"
lambda_zip_path      = "lambda.zip"
hf_token             = "hf_your_token_here"
fb_page_token        = "your_facebook_page_token"
fb_page_id           = "your_facebook_page_id"

⚠️ Important: Add terraform.tfvars to .gitignore to protect your secrets!

4. Build the Lambda Package

Windows (PowerShell):

.\build.ps1

Linux/Mac (Bash):

chmod +x deploy.sh
./deploy.sh

5. Deploy with Terraform

terraform init
terraform plan
terraform apply

Type yes when prompted to confirm deployment.

📁 Project Structure

.
├── lambda/
│   ├── main.py              # Main Lambda handler
│   ├── text_generator.py    # HF text generation
│   └── image_generator.py   # HF image generation
├── main.tf                  # Terraform infrastructure
├── variables.tf             # Terraform variables
├── outputs.tf               # Terraform outputs
├── requirements.txt         # Python dependencies
├── build.ps1               # Windows build script
├── build.sh                # Linux/Mac build script
└── README.md               # This file

🔧 Configuration

Lambda Function Settings

• Runtime: Python 3.13
• Memory: 512 MB (adjustable in main.tf)
• Timeout: 300 seconds (5 minutes)
• Schedule: Daily at 2 PM UTC (configurable)

Modify Schedule

Edit the cron expression in main.tf:

resource "aws_cloudwatch_event_rule" "daily_trigger" {
  schedule_expression = "cron(0 14 * * ? *)"  # 2 PM UTC daily
}

Cron format: cron(Minutes Hours Day-of-month Month Day-of-week Year)

Examples:

• Every hour: cron(0 * * * ? *)
• Every day at 9 AM UTC: cron(0 9 * * ? *)
• Every Monday at 10 AM UTC: cron(0 10 ? * MON *)

🧪 Testing

Test Lambda Function Locally

# In lambda/main.py, add at the bottom:
if __name__ == "__main__":
    import os
    os.environ['HF_TOKEN'] = 'your_token'
    os.environ['FB_PAGE_TOKEN'] = 'your_token'
    os.environ['FB_PAGE_ID'] = 'your_page_id'

    lambda_handler({}, None)

Run:

cd lambda
python main.py

Test in AWS Console

1. Go to AWS Lambda Console
2. Find your function
3. Click "Test" tab
4. Create a test event (empty {} is fine)
5. Click "Test"

📊 Monitoring

View Logs

# Using AWS CLI
aws logs tail /aws/lambda/ai-content-auto-poster --follow

Or view in AWS Console:

1. CloudWatch → Log Groups
2. Find /aws/lambda/ai-content-auto-poster

Check Metrics

• AWS Lambda Console → Your function → Monitor tab
• CloudWatch → Metrics → Lambda

🛠️ Troubleshooting

Package Too Large Error

If you get "Request must be smaller than 70MB" error:

zipped Size Too Large

If zipped size exceeds, you may need to:

1. Remove unnecessary dependencies
2. Use Lambda Layers
3. Switch to Container Image deployment

API Errors

Hugging Face API:

• Check token validity
• Verify model availability
• Check API rate limits

Facebook API:

• Verify Page Access Token hasn't expired
• Check page permissions
• Ensure Page ID is correct

🔐 Security Best Practices

1. Never commit secrets: Keep terraform.tfvars in .gitignore
2. Use AWS Secrets Manager: For production, store secrets in AWS Secrets Manager
3. Rotate tokens: Regularly rotate API tokens
4. Least privilege: IAM role has minimal required permissions

💰 Cost Estimation

AWS Costs (approximate):

• Lambda: ~$0.20/month (1 execution/day)
• EventBridge: Free tier
• CloudWatch Logs: ~$0.50/month

Total: ~$1/month

Hugging Face: Free tier includes limited API calls

Facebook: Free

🔄 Updates & Maintenance

Update Lambda Code

# Modify code in lambda/
.\build.ps1
terraform apply

Update Dependencies

# Edit requirements.txt
# Rebuild package
.\build.ps1
terraform apply

Destroy Infrastructure

terraform destroy

Type yes to confirm deletion of all resources.

🤝 Contributing

Contributions welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

📞 Support

• Issues: Open an issue on GitHub
• AWS Documentation: https://docs.aws.amazon.com/lambda/
• Hugging Face Docs: https://huggingface.co/docs
• Terraform Docs: https://registry.terraform.io/providers/hashicorp/aws/

🙏 Acknowledgments

• Hugging Face for API access
• AWS for serverless infrastructure
• Facebook for Graph API

---

Happy Posting! 🚀