Core runtime
| Variable | Description |
|---|---|
VITRUVIAN_PROFILE | Active profile selector |
VITRUVIAN_MODEL_PROVIDER | openai, anthropic, gemini |
VITRUVIAN_MODEL_NAME | Model override |
VITRUVIAN_MODEL_MAX_TOKENS | Anthropic max tokens override |
VITRUVIAN_WORKING_DIRECTORY | Workspace for file operations |
VITRUVIAN_MEMORY_CONNECTION_STRING | Durable memory/audit store |
Provider credentials
| Variable | Used when |
|---|---|
OPENAI_API_KEY | Provider is OpenAI |
ANTHROPIC_API_KEY | Provider is Anthropic |
GEMINI_API_KEY | Provider is Gemini |
Discord mode
| Variable | Description |
|---|---|
DISCORD_BOT_TOKEN | Bot authentication token |
DISCORD_CHANNEL_ID | Channel to poll |
DISCORD_POLL_INTERVAL_SECONDS | Poll frequency |
DISCORD_MESSAGE_LIMIT | Max messages fetched per cycle |
WebSocket mode
| Variable | Description |
|---|---|
VITRUVIAN_WEBSOCKET_URL | Listener endpoint |
VITRUVIAN_WEBSOCKET_PUBLIC_URL | Public URL shown to users |
VITRUVIAN_WEBSOCKET_DOMAIN | Domain tag for incoming requests |
Operations and posture
| Variable | Description |
|---|---|
VITRUVIAN_SECRET_PROVIDER | External secret provider integration |