Skip to main content

Frequently Asked Questions

Find answers to the most common questions about using Luz de Arcanos.
You can perform 5 consultations per day. This limit resets every 24 hours from your first consultation of the day. The limit is tracked using your browser’s localStorage to ensure fair usage of the AI service.If you’ve reached your daily limit, you’ll see a message: “Alcanzaste el límite de 5 consultas por día. Volvé mañana para seguir explorando.”
Yes, Luz de Arcanos is completely free and open source. The 5 consultations per day limit helps manage API costs for the Google Gemini AI service that powers the readings.The project is available on GitHub and you can even run your own instance with your own Gemini API key for unlimited consultations.
You can ask about almost any area of your life:
  • Career and professional decisions
  • Relationships and love
  • Personal growth and self-discovery
  • Financial decisions
  • Creative projects
  • Life direction and purpose
Health questions are not supported. If you ask about health, medical conditions, diagnoses, or pregnancy, Seraphina will politely decline and recommend consulting a medical professional instead. This is a safety feature built into the AI prompt.
Your question can be up to 500 characters long, so feel free to provide context and details.
Luz de Arcanos provides interpretive guidance rather than predictions. The AI (Seraphina) analyzes your question in context with three tarot cards drawn for past, present, and future positions.The readings are designed to be:
  • Reflective: Help you think about your situation from new angles
  • Practical: Offer actionable advice you can apply immediately
  • Empathetic: Delivered in a warm, conversational tone
Like traditional tarot, the value comes from the insights and perspectives the reading provides, not from literal prediction of the future.
No. Luz de Arcanos has built-in restrictions that prevent it from addressing health-related questions. This includes:
  • Medical conditions or diagnoses
  • Physical or mental health concerns
  • Pregnancy questions
  • Treatment decisions
If you ask a health-related question, Seraphina will respond with a polite message recommending you consult a qualified medical professional instead.This safety feature is implemented in the AI prompt at src/actions/index.ts:73.
Luz de Arcanos is built with modern web technologies:
  • Astro 5: Framework with SSR and Server Actions
  • Google Gemini API: AI models for generating personalized readings
  • TypeScript: Type-safe code throughout
  • Pure CSS: 3D card flip animations using transform-style: preserve-3d
  • Vercel: Deployment and hosting
The entire application is open source and available on GitHub.
When you submit a consultation, here’s what happens:
  1. Card Selection: Three random tarot cards are drawn from the 78-card deck (Major and Minor Arcana), each with a 50% chance of being reversed
  2. AI Processing: Your name, question, and the three cards (with their meanings and keywords) are sent to Google Gemini AI
  3. Model Fallback: The system tries multiple Gemini models in sequence (gemini-2.5-flash, gemini-2.0-flash, gemini-2.0-flash-lite) to handle API quota limits
  4. Reading Generation: The AI acts as “Seraphina,” a professional tarotist, and creates a personalized reading (max 250 words) covering:
    • Greeting and context
    • Past: What brought you here
    • Present: What’s happening now
    • Future: Where things are heading
    • Practical advice you can apply
  5. Fallback Response: If all AI models fail, a basic templated reading is provided so you always get a response
The entire process is implemented in src/actions/index.ts:28-97.
“Seraphina” is the persona name given to the AI in the system prompt. The AI is instructed to act as a professional, empathetic tarotist with a warm but direct and practical style.You’ll see references to Seraphina in:
  • Loading message: “Seraphina consulta las cartas para ti…”
  • Reading interpretation: “Seraphina está interpretando las cartas…”
This persona helps create a consistent, friendly tone that feels more personal than generic AI responses.
Have a question that isn’t answered here? Check the Support page for troubleshooting help or visit the GitHub repository to open an issue.

Build docs developers (and LLMs) love