Automatically send your starred Pocket Casts episodes to Snipcast.io for AI-powered summarization via AWS Lambda.

This AWS Lambda function:

• Authenticates with Pocket Casts and retrieves unplayed starred episodes
• Sends episode URLs to Snipcast API for summarization
• Automatically removes stars and archives processed episodes
• Returns detailed results including success/failure counts

• Python 3.12+
• AWS account with appropriate permissions
• Terraform installed (for deployment)
• Pocket Casts account
• Snipcast.io account with API key

1. Clone and setup:

git clone https://github.com/essoen/pocketcast-to-snipcast
cd pocketcast-to-snipcast
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

2. Configure environment:

cp .env.example .env
# Edit .env with your credentials

3. Test locally:

source venv/bin/activate
python lambda_function.py

See detailed instructions in CLAUDE.md.

Quick start:

cd terraform
cp terraform.tfvars.example terraform.tfvars
# Edit terraform.tfvars with your credentials
terraform init
terraform plan
terraform apply

1. Lambda Function (lambda_function.py) - Main entry point and processing logic
2. Pocket Casts Client (pocketcasts_client.py) - Custom API client
3. Terraform Infrastructure (terraform/) - AWS deployment configuration

1. EventBridge triggers Lambda on schedule (default: every 2 hours)
2. Lambda authenticates with Pocket Casts
3. Fetches starred episodes, filters for unplayed (playingStatus == 1)
4. For each episode:
   • Constructs share URL: https://pca.st/episode/{episode_uuid}
   • Posts to Snipcast API
   • If successful: unstars and archives episode
5. Returns JSON summary with counts and episode details

pocketcast-to-snipcast/
├── lambda_function.py          # Main Lambda handler
├── pocketcasts_client.py       # Pocket Casts API client
├── requirements.txt            # Python dependencies
├── .env.example                # Environment variables template
├── terraform/                  # Infrastructure as Code
│   ├── main.tf
│   ├── variables.tf
│   ├── outputs.tf
│   └── terraform.tfvars.example
├── lambda-deployment/          # Deployment package (auto-generated)
├── CLAUDE.md                   # Detailed architecture guide
└── README.md                   # This file

Required:

• POCKET_CASTS_EMAIL: Pocket Casts account email
• POCKET_CASTS_PASSWORD: Pocket Casts account password
• SNIPCAST_API_KEY: Snipcast.io API key

Edit terraform/terraform.tfvars:

schedule_expression = "rate(1 hour)"   # or "cron(0 12 * * ? *)"
schedule_rate = "1 hour"

Then: terraform apply

• Verify credentials in terraform.tfvars
• Ensure using email/password (not third-party auth)

• Install dependencies: cd lambda-deployment && pip install -r ../requirements.txt -t .
• Redeploy: terraform apply

• Check CloudWatch logs for unstar/archive operation status
• Lambda uses Pocket Casts playing status to avoid reprocessing

# Real-time logs
aws logs tail /aws/lambda/pocketcast-to-snipcast --region eu-north-1 --follow

# Get CloudWatch URL
terraform output cloudwatch_logs_url

aws lambda invoke \
  --function-name pocketcast-to-snipcast \
  --region eu-north-1 \
  response.json

cat response.json | jq

• All credentials passed as environment variables
• Terraform variables marked as sensitive = true
• Never commit terraform.tfstate or terraform.tfvars
• IAM role follows principle of least privilege

MIT License