Docker

Package sportsclaw — the engine and the sports data layer — into a single container, ready to run a bot anywhere.

Build and run

npm run docker:build      # build the image
npm run docker:run        # run it with your .env file

The image bundles both the Node engine and the Python data layer, so the container is self-contained.

Configuration

Pass your credentials and bot tokens with an env file:

docker run --rm --env-file .env sportsclaw

Variable	Purpose
ANTHROPIC_API_KEY / OPENAI_API_KEY / GEMINI_API_KEY	Your model provider
DISCORD_BOT_TOKEN	Run a Discord bot
TELEGRAM_BOT_TOKEN	Run a Telegram bot

Persisting settings

sportsclaw stores config, installed sports, and subscriptions under ~/.sportsclaw. Mount a volume there if you want that state to survive container restarts:

docker run --rm --env-file .env -v sportsclaw-data:/root/.sportsclaw sportsclaw