Quickstart

Requirements

• Docker with the Compose plugin
• A reachable Radarr, Sonarr, Lidarr, or Whisparr instance
• The API key from Settings → General in that arr application

1. Download the deployment files

mkdir warrden && cd warrden
curl -LO https://raw.githubusercontent.com/neurekadev/warrden/main/compose.yaml
curl -Lo config.yaml https://raw.githubusercontent.com/neurekadev/warrden/main/config.example.yaml
curl -Lo .env https://raw.githubusercontent.com/neurekadev/warrden/main/.env.example

The supplied Compose file uses ghcr.io/neurekadev/warrden:5, persists state in a named volume, and mounts config.yaml at /app/data/config.yaml.

2. Configure an instance

Open config.yaml, find the block for your arr application, and set at least these fields:

instances:
  - type: sonarr
    enabled: true
    name: Series
    url: http://sonarr:8989
    apiVersion: v3
    apiKey: YOUR_SONARR_API_KEY

Keep the job sections from the example file, then adjust their schedules and limits when you are ready. Instance names must be unique.

If the arr application is outside the Compose network, use a URL that is reachable from inside the wArrden container.

3. Preview actions

Set dry-run mode in .env before the first start:

DRY_RUN=true

Dry-run mode leaves queues and searches unchanged while showing what cleanup rules would match.

4. Start wArrden

docker compose up -d
docker compose logs -f warrden

Confirm the startup summary lists the expected instance and scheduled jobs. When the output looks correct, change DRY_RUN=false and recreate the container:

docker compose up -d --force-recreate

Updating

docker compose pull
docker compose up -d

The named data volume keeps cooldown history across image updates.