Skip to content

Configuration Reference

Jump to bottom
mrveiss edited this page Jul 22, 2026 · 2 revisions

Configuration Reference

All configuration is done through environment variables in the .env file. Copy .env.example to .env before first run.

cp .env.example .env
nano .env

Core Settings

Variable Description Default Required
AUTOBOT_SECRET_KEY JWT signing key — generate with openssl rand -hex 32 — Yes
AUTOBOT_ENV Runtime environment: development, production production No
AUTOBOT_DEBUG Enable debug mode and verbose logging false No
AUTOBOT_ADMIN_USERNAME Username for the first admin account seeded at startup admin No
AUTOBOT_ADMIN_PASSWORD Password for that first admin — change before deploying changeme Yes

Multi-user: These variables only seed the initial admin. AutoBot is a multi-user platform with role-based access control (RBAC) — additional users and their roles are created and managed in-app after first login, not through environment variables. Single-user mode is retired.

Database

Variable Description Default
POSTGRES_HOST PostgreSQL hostname autobot-database
POSTGRES_PORT PostgreSQL port 5432
POSTGRES_DB Database name autobot
POSTGRES_USER Database user autobot
POSTGRES_PASSWORD Database password autobot
DATABASE_URL Full connection string (overrides individual settings) —

Redis

Variable Description Default
REDIS_HOST Redis hostname redis
REDIS_PORT Redis port 6379
REDIS_PASSWORD Redis password (leave empty for no auth) —
REDIS_DB Redis database number 0

ChromaDB (Vector Database)

Variable Description Default
CHROMA_HOST ChromaDB hostname chromadb
CHROMA_PORT ChromaDB port 8001
CHROMA_COLLECTION Default collection name autobot

AI / LLM Settings

AutoBot routes all inference through an LLM gateway. You choose the backend: local inference via Ollama, an external provider, or a mix with fallback. The variables below configure local inference; provider selection and credentials are described under LLM Gateway & Providers.

Variable Description Default
OLLAMA_HOST Ollama server hostname autobot-ai-stack
OLLAMA_PORT Ollama port 11434
OLLAMA_MODEL Default model for chat llama3
OLLAMA_EMBED_MODEL Model used for embeddings nomic-embed-text
LLM_CONTEXT_WINDOW Max context tokens 4096
LLM_MAX_TOKENS Max tokens per response 2048
LLM_TEMPERATURE Model temperature (0.0–1.0) 0.7

LLM Gateway & Providers

In addition to local Ollama, the gateway can route requests to external providers such as claude_api and openrouter, with fallback between them. Provider selection and API credentials are configured through the gateway's provider settings.

  • Local only: leave providers unconfigured — inference stays on the host and nothing is sent externally.
  • External / hybrid: configure the desired provider and its API key to enable routing and fallback.

Exact provider variable names depend on your build — check .env.example and the in-app model/provider settings for the authoritative list. Your data stays on your infrastructure; a provider is contacted only if you configure one.

Supported Models

Any model supported by Ollama can be used. Popular choices:

Model Size Best For
llama3 8B General use, good balance
llama3:70b 70B Best quality (requires more RAM)
mistral 7B Fast, good for code
codellama 7B Code-focused tasks
nomic-embed-text — Embeddings (required for RAG)

Pull a model:

docker compose exec autobot-ai-stack ollama pull llama3

Auth & Security

Variable Description Default
JWT_EXPIRY_HOURS Token expiry in hours 24
JWT_REFRESH_EXPIRY_DAYS Refresh token expiry 7
CORS_ORIGINS Allowed CORS origins (comma-separated) http://localhost
SESSION_COOKIE_SECURE Require HTTPS for session cookies true

TLS / HTTPS

Variable Description Default
TLS_ENABLED Enable HTTPS in nginx false
TLS_CERT_PATH Path to TLS certificate —
TLS_KEY_PATH Path to TLS private key —
FORCE_HTTPS_REDIRECT Redirect HTTP → HTTPS false

Fleet (Ansible)

Variable Description Default
ANSIBLE_SSH_USER Default SSH user for fleet nodes autobot
ANSIBLE_SSH_KEY_PATH Path to SSH private key /etc/autobot/ssh/id_rsa
ANSIBLE_TIMEOUT SSH connection timeout (seconds) 30
FLEET_AGENT_PORT Port for AutoBot agent on nodes 8080

Storage

Variable Description Default
UPLOAD_MAX_SIZE_MB Maximum file upload size 100
UPLOAD_PATH Path for uploaded documents /data/uploads
KB_CHUNK_SIZE Knowledge base document chunk size (tokens) 512
KB_CHUNK_OVERLAP Chunk overlap for context continuity 64

Monitoring (Optional)

Variable Description Default
PROMETHEUS_ENABLED Enable Prometheus metrics endpoint false
METRICS_PORT Metrics endpoint port 9090
GRAFANA_ENABLED Enable Grafana dashboard false
GRAFANA_PORT Grafana port 3000

To enable monitoring:

PROMETHEUS_ENABLED=true
GRAFANA_ENABLED=true

Then start with:

docker compose --profile monitoring up -d

Logging

Variable Description Default
LOG_LEVEL Log level: debug, info, warning, error info
LOG_FORMAT json or text json
LOG_FILE Write logs to file (optional) —

Example Production .env 

# Core
AUTOBOT_SECRET_KEY=your-64-char-hex-string-here
AUTOBOT_ENV=production
AUTOBOT_ADMIN_PASSWORD=strong-password-here

# Database
POSTGRES_PASSWORD=strong-db-password
POSTGRES_USER=autobot
POSTGRES_DB=autobot

# AI
OLLAMA_MODEL=llama3
OLLAMA_EMBED_MODEL=nomic-embed-text

# TLS
TLS_ENABLED=true
TLS_CERT_PATH=/certs/fullchain.pem
TLS_KEY_PATH=/certs/privkey.pem
FORCE_HTTPS_REDIRECT=true

# Security
JWT_EXPIRY_HOURS=8
SESSION_COOKIE_SECURE=true

Clone this wiki locally