openarg_backend: Command for Claude Code

Create new Celery worker: $ARGUMENTS

Create a new worker with its task and dedicated queue. Follow these steps:

1. Create the tasks file

Create src/app/infrastructure/celery/tasks/{name}_tasks.py:

from src.app.infrastructure.celery.app import celery_app

@celery_app.task(
    name="openarg.{name}_task",
    bind=True,
    max_retries=3,
    default_retry_delay=60,
)
def {name}_task(self, param1: str, param2: str | None = None):
    """Description of what the worker does."""
    try:
        # 1. Setup (resolve dependencies, create session)
        # 2. Execute logic
        # 3. Return result or dispatch next task
        pass
    except Exception as exc:
        self.retry(exc=exc)

Reference: src/app/infrastructure/celery/tasks/scraper_tasks.py (scrape + dispatch), src/app/infrastructure/celery/tasks/analyst_tasks.py (multi-step pipeline).

2. Configure queue routing

Edit src/app/infrastructure/celery/app.py, add to the task_routes dict:

task_routes = {
    ...
    "openarg.{name}_task": {"queue": "{name}"},
}

3. Add Make command

Edit the Makefile, add:

workers.{name}:
	celery -A src.app.infrastructure.celery.app worker -Q {name} -c {concurrency} --loglevel=info -n {name}@%h

Reference: look for workers.scraper, workers.collector in the Makefile.

4. Add Docker service

Edit docker-compose.yaml, add:

  worker-{name}:
    build:
      context: .
      dockerfile: docker/worker.Dockerfile
    command: celery -A src.app.infrastructure.celery.app worker -Q {name} -c {concurrency} --loglevel=info
    environment:
      - CELERY_BROKER_URL=redis://redis:6379/0
      - CELERY_RESULT_BACKEND=redis://redis:6379/1
      - DATABASE_URL=postgresql+psycopg://openarg:openarg@postgres:5432/openarg
    depends_on:
      - redis
      - postgres

Reference: look for worker-scraper, worker-analyst in docker-compose.yaml.

5. Dispatch the task

From an endpoint or from another worker:

from src.app.infrastructure.celery.tasks.{name}_tasks import {name}_task

{name}_task.delay(param1, param2)
# or with explicit routing:
{name}_task.apply_async(args=[param1], queue="{name}")