Configuration Reference

Server Configuration

Configuration is loaded from server/config/sys.config or environment variables.

Core Settings

Setting	Default	Description
port	8001	HTTP listener port
tls_enabled	false	Enable HTTPS
tls_cert_file	—	TLS certificate path
tls_key_file	—	TLS private key path
auth_enabled	false	Require authentication
workspace_root	.	Workspace files root directory
session_dir	./sessions	Session storage directory

Inference Settings

Setting	Default	Description
inference_bin	./llama-server	llama.cpp server binary
inference_gpu_layers	99	GPU layers to offload
inference_ctx_size	4096	Context window size
inference_threads	4	CPU threads

Memory Settings

Setting	Default	Description
memory_pg_dsn	—	PostgreSQL DSN for vector storage
memory_embed_model	—	Embedding model path

ZMQ Settings

Setting	Default	Description
zmq_router_port	5555	ZMQ ROUTER bind port
zmq_pub_port	5556	ZMQ PUB bind port

Monitoring

Setting	Default	Description
prometheus_enabled	true	Expose /metrics endpoint
otel_enabled	false	OpenTelemetry tracing

CLI Configuration

Stored at ~/.zerg/config.json:

{
  "profiles": {
    "default": {
      "url": "https://api.nonsense.ws",
      "token": "your-api-token"
    },
    "staging": {
      "url": "https://staging.nonsense.ws",
      "token": "staging-token"
    }
  }
}

Luna Agent Configuration

Luna uses ~/.luna/config.json:

{
  "provider": "anthropic",
  "model": "claude-sonnet-4-20250514",
  "permission_mode": "default",
  "theme": "default"
}