Environment Variables

Frona AI is configured through environment variables. All variables use the FRONA_ prefix and are organized by component.

General

Variable	Default	Description
FRONA_SERVER_DATA_DIR	data	Base data directory. Defaults for FRONA_CONFIG, database.path, storage.workspaces_path, and storage.files_path are derived from this path
FRONA_CONFIG	{data_dir}/config.yaml	Path to the YAML config file
FRONA_LOG_LEVEL	—	Log verbosity level
FRONA_LOG_CONFIG	—	Path to a custom log configuration file

Server

Variable	Default	Description
FRONA_SERVER_PORT	3001	Port the backend API listens on
FRONA_SERVER_BASE_URL	—	Public base URL of the server (used for callbacks)
FRONA_SERVER_MAX_BODY_SIZE_BYTES	104857600 (100 MB)	Maximum request body size
FRONA_SERVER_CORS_ORIGINS	—	Comma-separated list of allowed CORS origins
FRONA_SERVER_MAX_CONCURRENT_TASKS	10	Maximum concurrent tasks across all agents
FRONA_SERVER_SANDBOX_DISABLED	false	Disable CLI sandboxing (not recommended for production)

Authentication

Variable	Default	Description
FRONA_AUTH_ENCRYPTION_SECRET	—	Secret used to derive the AES-256 key that encrypts JWT signing keypairs at rest. Must be changed in production.
FRONA_AUTH_ACCESS_TOKEN_EXPIRY_SECS	900 (15 min)	Access token lifetime
FRONA_AUTH_REFRESH_TOKEN_EXPIRY_SECS	604800 (7 days)	Refresh token lifetime
FRONA_AUTH_PRESIGN_EXPIRY_SECS	86400 (24 hours)	Pre-signed URL lifetime

Change the encryption secret in production

FRONA_AUTH_ENCRYPTION_SECRET is used to derive an AES-256 encryption key (via SHA-256) that protects the JWT signing keypairs stored in the database. It is not used directly for JWT signing — instead it encrypts the private keys that do the signing.

A built-in default is provided for local development, but you must set your own value in production. If the default is left in place and database files are ever exposed (backup leak, file traversal, shared host), an attacker could decrypt the signing keypairs and forge authentication tokens for any user.

Generate a strong random secret:

openssl rand -base64 32

SSO (OpenID Connect)

Variable	Default	Description
FRONA_SSO_ENABLED	false	Enable OIDC authentication
FRONA_SSO_AUTHORITY	—	OpenID Connect authority URL
FRONA_SSO_CLIENT_ID	—	OIDC client ID
FRONA_SSO_CLIENT_SECRET	—	OIDC client secret
FRONA_SSO_SCOPES	email profile offline_access	OpenID scopes to request
FRONA_SSO_ONLY	false	Force SSO-only authentication, disables local login
FRONA_SSO_SIGNUPS_MATCH_EMAIL	true	Match SSO signups to existing accounts by email
FRONA_SSO_ALLOW_UNKNOWN_EMAIL_VERIFICATION	false	Accept unverified emails from the IdP
FRONA_SSO_CLIENT_CACHE_EXPIRATION	0	Client metadata cache expiration in seconds

Database

Variable	Default	Description
FRONA_DATABASE_PATH	data/db	Path to the SurrealDB data directory

Browser

Variable	Default	Description
FRONA_BROWSER_WS_URL	—	WebSocket URL of the Browserless instance
FRONA_BROWSER_PROFILES_PATH	/profiles	Path for storing browser profiles
FRONA_BROWSER_CONNECTION_TIMEOUT_MS	30000	Timeout for connecting to the browser

Search

Variable	Default	Description
FRONA_SEARCH_PROVIDER	—	Search provider: searxng, tavily, or brave
FRONA_SEARCH_SEARXNG_BASE_URL	—	Base URL of the SearXNG instance

Storage

Variable	Default	Description
FRONA_STORAGE_WORKSPACES_PATH	data/workspaces	Path for workspace file storage
FRONA_STORAGE_FILES_PATH	data/files	Path for file uploads
FRONA_STORAGE_SHARED_CONFIG_DIR	resources	Path for shared prompts and agent configs

Scheduler

Variable	Default	Description
FRONA_SCHEDULER_POLL_SECS	60	How often the scheduler checks for due tasks
FRONA_SCHEDULER_SPACE_COMPACTION_SECS	3600 (1 hour)	Interval for space context compaction
FRONA_SCHEDULER_INSIGHT_COMPACTION_SECS	7200 (2 hours)	Interval for memory insight compaction

Inference

Variable	Default	Description
FRONA_INFERENCE_MAX_TOOL_TURNS	200	Maximum tool call iterations per response
FRONA_INFERENCE_DEFAULT_MAX_TOKENS	8192	Default max tokens for LLM responses
FRONA_INFERENCE_COMPACTION_TRIGGER_PCT	80	Context usage percentage that triggers compaction
FRONA_INFERENCE_HISTORY_TRUNCATION_PCT	90	Context usage percentage that triggers truncation

Voice (Twilio)

Variable	Default	Description
FRONA_VOICE_PROVIDER	—	Voice provider. Currently only twilio is supported
FRONA_VOICE_TWILIO_ACCOUNT_SID	—	Twilio account SID
FRONA_VOICE_TWILIO_AUTH_TOKEN	—	Twilio auth token
FRONA_VOICE_TWILIO_FROM_NUMBER	—	Twilio phone number for outbound calls (E.164 format)
FRONA_VOICE_TWILIO_VOICE_ID	—	Twilio voice ID for text-to-speech
FRONA_VOICE_TWILIO_SPEECH_MODEL	—	Twilio speech recognition model
FRONA_VOICE_CALLBACK_BASE_URL	—	Public URL for Twilio callbacks (overrides server.base_url)

LLM Provider

Variable	Default	Description
ANTHROPIC_API_KEY	—	API key for Anthropic (Claude)
OPENAI_API_KEY	—	API key for OpenAI
GROQ_API_KEY	—	API key for Groq
OPENROUTER_API_KEY	—	API key for OpenRouter
DEEPSEEK_API_KEY	—	API key for DeepSeek
GEMINI_API_KEY	—	API key for Google Gemini
COHERE_API_KEY	—	API key for Cohere
MISTRAL_API_KEY	—	API key for Mistral
PERPLEXITY_API_KEY	—	API key for Perplexity
TOGETHER_API_KEY	—	API key for Together AI
XAI_API_KEY	—	API key for xAI (Grok)
HYPERBOLIC_API_KEY	—	API key for Hyperbolic
MOONSHOT_API_KEY	—	API key for Moonshot
MIRA_API_KEY	—	API key for Mira
GALADRIEL_API_KEY	—	API key for Galadriel
HUGGINGFACE_API_KEY	—	API key for Hugging Face
OLLAMA_API_BASE_URL	—	Base URL for Ollama instance

Model groups

Model groups are configured in the application config file rather than environment variables. Each group defines:

• main: the primary model identifier
• fallbacks: a list of fallback models
• max_tokens: token limit override
• temperature: temperature override
• context_window: context window size
• retry: exponential backoff settings for retries