Skip to content

quickstart.md

Latest commit

 

History

History
62 lines (41 loc) · 1.87 KB

quickstart.md

File metadata and controls

62 lines (41 loc) · 1.87 KB

Quick Start

Docker (recommended)

curl -O https://raw.githubusercontent.com/techbutton/arrmate/main/docker-compose.prod.yml
curl -O https://raw.githubusercontent.com/techbutton/arrmate/main/.env.example
cp .env.example .env

Open .env and fill in at minimum:

  • LLM_PROVIDER and the corresponding model/key settings
  • URLs and API keys for any services you want to connect
docker compose -f docker-compose.prod.yml up -d

Then open http://localhost:8000.

If you're using Ollama, pull a model:

docker compose -f docker-compose.prod.yml exec ollama ollama pull qwen2.5:7b

qwen2.5:7b works well for tool-calling. Larger models will be more accurate with complex commands.

Local Python

git clone https://github.com/techbutton/arrmate.git
cd arrmate
python3 -m venv venv && source venv/bin/activate
pip install -e .
cp .env.example .env
# edit .env
python -m arrmate.interfaces.api.app

Finding API keys

Sonarr / Radarr / Lidarr / Bazarr: Settings → General → Security → API Key

Plex token: Sign in at app.plex.tv, open any item, click ··· → Get Info → View XML — the token is in the URL as X-Plex-Token.

AudioBookshelf: Settings → Security → API Tokens

Troubleshooting

Services show as unavailable — verify the URL is reachable from inside the container. Use Docker service names (e.g. http://sonarr:8989), not localhost, when services are on the same Docker network.

Commands aren't parsed correctly — your Ollama model may not support tool-calling reliably. qwen2.5:7b or llama3.1 are good options. For OpenAI/Anthropic the defaults work fine.

Can't reach Sonarr/Radarr — test from inside the container:

docker exec arrmate curl http://sonarr:8989/api/v3/system/status?apikey=YOUR_KEY

Locked out — delete auth.json from your data volume and restart.