Skip to content

Latest commit

 

History

History
112 lines (79 loc) · 6.53 KB

File metadata and controls

112 lines (79 loc) · 6.53 KB

Settings

Atlantis provides configurable settings accessible via the /settings page.

Appearance

Setting Options Default Description
Theme Light, Dark, System System Application color scheme

Editor

Setting Options Default Description
Auto-save On/Off On Automatically save changes
Auto-save delay 1s, 2s, 5s, 10s 2s Debounce time before auto-saving

AI Settings

Setting Description
API Key Your OpenAI or Gemini API key (stored locally or via env var)
Provider Auto-detect, OpenAI-compatible, or Gemini
Model Read-only display of the model in use

Supported Models

  • OpenAI: gpt-4o-mini
  • Gemini: gemini-3.1-flash-lite-preview (configurable via GEMINI_MODEL env var)

Note: When AI_API_KEY environment variable is set, the key cannot be modified through the UI and is shown as "read-only".

Advanced

Setting Range Default Description
Max checkpoints 5-50 15 Maximum version history per diagram
Default export format SVG, PNG, PDF SVG Preferred diagram download format
Export scale 1x, 2x, 3x 2x PNG/PDF resolution multiplier

Data

Action Description
Backup Download all diagrams and notes as JSON
Restore Upload a previously downloaded backup (validated with strict schema)

Danger Zone

Action Description
Wipe Database Permanently delete all data (requires server-generated confirmation code)

Environment Variables

Settings can also be configured via environment variables. These are especially useful for Docker deployments.

Complete Variable Reference

Variable Default Description
PORT 3000 Port the application listens on
NODE_ENV development (local), production (Docker/next start) Environment mode (development, production)
PRISMA_PROVIDER sqlite Database provider (sqlite, postgresql, mysql)
DATABASE_URL file:./data/atlantis.db Database connection string (DB_CONNECTION alias supported)
REDIS_URL (none) Redis connection (e.g., redis://redis:6379)
AI_API_KEY (none) AI API key for OpenAI/Gemini features
GEMINI_MODEL gemini-3.1-flash-lite-preview Gemini model to use
GEMINI_API_VERSION v1 or v1beta (auto-detected) Gemini API version (defaults to v1beta for -preview models)
ENABLE_API_ACCESS false Enable REST API endpoints and /docs (requires CSRF cookie/header for writes)
PRISMA_AUTO_APPLY true (dev), false (prod) Dev bootstrap only: auto-run prisma db push during npm run dev/npm run build (Docker entrypoint/npm run start always runs prisma db push)
PRISMA_SKIP_AUTOPUSH false Dev bootstrap only: skip automatic db push/seed in scripts/bootstrap.js (no effect on npm run start or Docker)

Docker Compose Usage

Set variables inline when starting services:

PORT=8080 ENABLE_API_ACCESS=true docker compose up -d

Or create a .env file in your project root:

PORT=3000
ENABLE_API_ACCESS=true
AI_API_KEY=your-api-key-here
REDIS_URL=redis://redis:6379

Then run:

docker compose up -d

Security Notes

  • AI API Key: Using AI_API_KEY env var is more secure than storing in the database. When set, the key is read-only in the UI.
  • Wipe Confirmation: The confirmation code is generated server-side with a 5-minute TTL and timing-safe validation.
  • Backup Restore: Imported backups are validated against a strict Zod schema before processing.

Related Documentation