SLEEPARR

Personal project in BETA! I built this for my own homelab using the tools I use daily: Zed, Claude, and OpenCode. AI is part of my regular workflow the same way Stack Overflow and docs always have been. If you spot any issues, chuck in a ticket and I'll get to it.

A self-hosted dashboard for searching missing media across your arr stack.

Currently supports Sonarr. Radarr support coming soon.

Quick start

# 1. Copy the env file and set your JWT secret
cp .env.example .env
# Edit .env — generate a secret with:
# openssl rand -hex 32

# 2. Start
docker compose up -d

# 3. Open http://localhost:8080 and create your admin account

docker-compose.yml

services:
  sleeparr:
    image: ghcr.io/codevski/sleeparr:latest
    container_name: sleeparr
    restart: unless-stopped
    ports:
      - "${PORT:-8080}:8080"
    volumes:
      - ./config:/config
    environment:
      - JWT_SECRET=${JWT_SECRET:-change-me-generate-with-openssl-rand-hex-32}
      - GIN_MODE=${GIN_MODE:-release}
      - PUID=${PUID:-1000}
      - PGID=${PGID:-1000}
      - TZ=${TZ:-America/New_York}
    env_file:
      - .env

Configuration

All configuration is stored in /config/config.json and managed through the UI. Environment variables are for startup only.

Variable	Required	Default	Description
JWT_SECRET	Yes	—	Secret for signing JWT tokens. Generate with openssl rand -hex 32
PORT	No	8080	Port to listen on
DATA_DIR	No	/config	Path to store config.json
GIN_MODE	No	release	Set to debug to enable verbose request logging

File permissions (NAS / homelab)

If file ownership matters on your setup, set PUID and PGID to match your user:

id
# uid=1000(youruser) gid=1000(yourgroup)

environment:
  - PUID=1000
  - PGID=1000
  - TZ=Australia/Melbourne

This ensures /config files are owned by your user rather than root.

Development

# Backend (with hot reload)
air

# Frontend (separate terminal)
cd frontend && bun run dev

See bruno/sleeparr/ for the API collection.

License

MIT