Skip to content

Configuration

AIMQ can be configured through environment variables and requires proper Supabase queue setup.

Supabase Queue Setup

  1. Enable Queue Integration:
  2. Go to your Supabase project dashboard
  3. Navigate to Database → Extensions
  4. Enable the "pg_net" and "pg_cron" extensions if not already enabled
  5. Navigate to Database → Queues (Beta)
  6. Click "Enable Queue"

  7. Make sure to enable "Expose Queues via PostgREST"

  8. Create Queues:

  9. In the Queues interface, click "Create a new queue"
  10. Give your queue a name (this will be referenced in your @worker.task decorators)
  11. Configure queue settings as needed

For more details, see the Supabase Queue Documentation.

Environment Variables

The following environment variables are supported:

# Required Supabase Configuration
SUPABASE_URL=your-project-url
SUPABASE_KEY=your-service-role-key  # Must be service role key, not anon key

# Worker Configuration (Optional)
WORKER_NAME=my-worker  # Default: 'peon'
WORKER_LOG_LEVEL=info  # Default: 'info'
WORKER_IDLE_WAIT=10.0  # Default: 10.0 seconds

# LangChain Configuration (Optional)
LANGCHAIN_TRACING_V2=true  # Enable LangChain tracing
LANGCHAIN_ENDPOINT=https://api.smith.langchain.com
LANGCHAIN_API_KEY=your-langchain-api-key
LANGCHAIN_PROJECT=your-project-name

# OpenAI Configuration (If using OpenAI)
OPENAI_API_KEY=your-openai-api-key

Configuration File

You can create a .env file in your project root:

# .env
SUPABASE_URL=your-project-url
SUPABASE_KEY=your-service-role-key
WORKER_NAME=my-worker

Using Poetry

Since this project uses Poetry for dependency management, you can:

  1. Install dependencies: 

    poetry install

  2. Run with environment variables: 

    poetry run python -m aimq.worker

Or use your .env file: 

poetry run python -m aimq.worker

Configuration in Code

Access configuration in your code:

from aimq.config import config

# Access configuration values
supabase_url = config.supabase_url
worker_name = config.worker_name

# Create a worker with custom configuration
from aimq import Worker

worker = Worker(
    name="custom-worker",
    log_level="debug",
    idle_wait=5.0
)

Next Steps