Joel Discord Bot

A Discord bot with AI-powered responses and message tracking.

Quick Start

# Install dependencies
bun install

# Set environment variables
cp .env.example .env
# Edit .env with your tokens

# Run in development mode
bun run dev

# Run in production
bun run start

Project Structure

See src/README.md for detailed architecture documentation.

src/
├── core/       # Bot client, config, logging
├── commands/   # Slash commands
├── events/     # Discord event handlers
├── features/   # Feature modules (Joel AI, message logging)
├── services/   # External services (AI providers)
├── database/   # Database schema and repositories
└── utils/      # Shared utilities

Environment Variables

Variable	Description
DISCORD_TOKEN	Discord bot token
DISCORD_CLIENT_ID	Discord application client ID
DISCORD_CLIENT_SECRET	Discord application client secret
OPENROUTER_API_KEY	OpenRouter API key for AI
FAL_API_KEY or FAL_KEY	Fal API key for image generation
AI_CLASSIFICATION_FALLBACK_MODELS	Comma-separated fallback model IDs for classification requests
KLIPY_API_KEY	Klipy API key for GIF search (optional)
ELEVENLABS_API_KEY	ElevenLabs API key for voiceover
ELEVENLABS_VOICE_ID	Default ElevenLabs voice ID (optional)
ELEVENLABS_MODEL	ElevenLabs model ID (default: eleven_multilingual_v2)
DATABASE_PATH	SQLite file path (default: ./data/db.sqlite3)
WEB_PORT	Port for web dashboard (default: 3000)
WEB_BASE_URL	Base URL for web dashboard
SESSION_SECRET	Secret for session encryption

Data Persistence

• Bot options and other runtime state are stored in SQLite.
• Keep DATABASE_PATH on persistent storage so updates/redeploys do not reset settings.
• Backward compatibility: if DATABASE_PATH is unset and src/database/db.sqlite3 exists, it is used automatically.
• Recommended one-time migration:

mkdir -p data && cp src/database/db.sqlite3 data/db.sqlite3

• Docker example:

docker run -d \
	--name joel-discord \
	-v joel_data:/data \
	--env-file .env \
	your-image:latest

Docker

Build the image:

docker build -t ghcr.io/your-org/joel-bot:latest .

Run it locally:

docker run -d \
  --name joel-bot \
  -p 3000:3000 \
  -v joel_data:/data \
  --env-file .env \
  ghcr.io/your-org/joel-bot:latest

The container now starts in production mode and does not enable the CSS file watcher.

k3s

The repo includes starter manifests in k8s/deployment.yaml, k8s/secret.example.yaml, and k8s/ingress.example.yaml.

Typical flow:

# Build and push the image
docker build -t ghcr.io/your-org/joel-bot:latest .
docker push ghcr.io/your-org/joel-bot:latest

# Create your secret manifest from the example and fill in real values
cp k8s/secret.example.yaml k8s/secret.yaml

# Apply namespace, PVC, deployment, and service
kubectl apply -f k8s/deployment.yaml
kubectl apply -f k8s/secret.yaml

# Optional if you want the web UI exposed through Traefik
kubectl apply -f k8s/ingress.example.yaml

Notes:

• Update the image in k8s/deployment.yaml to your registry path.
• Set WEB_BASE_URL in the secret to the public HTTPS URL you expose through ingress.
• The deployment mounts /data on a PVC and stores SQLite at /data/db.sqlite3.
• k3s usually provisions the PVC automatically via the default local-path storage class.

Scripts

Script	Description
bun run dev	Development with hot reload
bun run start	Production start
bun run build	Build for production
bun run db:generate	Generate DB migrations
bun run db:migrate	Run DB migrations
bun run db:studio	Open Drizzle Studio
bun run typecheck	Type check without emit