VEIL & ARCANA / AI READING / 02
TAROT
CHOOSEREVEAL
LISTEN
REAL PRODUCT / VEIL & ARCANA / LIVEChoose the cards. Read the pattern between them.
PHYSICAL SELECTION.
STRUCTURED INTERPRETATION.
TAROT is a Next.js and TypeScript experience with a React interface and a Three.js card table. The browser owns the deck, shuffle, selected cards and orientations; the server rebuilds trusted card meaning before requesting an AI interpretation.
THE EXPERIENCE / 02
TAROT is the Sun X Project page for Veil & Arcana, an English-language, AI-assisted Major Arcana reading experience. Visitors choose a spread, deck and card back, select their own cards on a cinematic 3D table, then receive a structured OpenAI interpretation with optional OpenAI or ElevenLabs narration.
01Four one-, three- and five-card reading spreads
02Two available Major Arcana decks and five card-back designs
03Local shuffle, card selection and upright or reversed orientation
04Server-validated OpenAI interpretation with optional generated narration
- 01CHOOSE
Select a one-, three- or five-card spread, then choose the physical deck and card-back design.
- 02SHUFFLE
Approach the cinematic table, shuffle locally and pull every card directly from the 3D fan.
- 03REVEAL
Let the cards turn in spread order while the server validates their identities, positions and orientations.
- 04LISTEN
Read the structured OpenAI interpretation or generate narration through the selected voice provider.
THE CARDS
STAY IN YOUR HANDS.
Veil & Arcana is an English-language Major Arcana experience built around a direct ritual: choose a spread, choose the physical deck and card back, then shuffle and select the cards yourself on a cinematic table.
The AI enters only after the selection is complete. It interprets the validated cards and positions, shows its model and response reference, and can narrate the finished reading through a selectable voice provider.
ENTER.
SHUFFLE.
CHOOSE.
The scene is deliberately quiet before the first card moves. A layered reader portrait responds subtly to the pointer while the Three.js table keeps shuffling, selection, dealing and reveal inside one continuous space.
The reading begins with a physical action, not a generated answer.
The selected spread waits on the table until the visitor chooses to shuffle the deck.
REAL PRODUCT CAPTURE / DESKTOP
A physical fan of card backs keeps the final draw under the visitor’s control.
LOCAL SELECTION / 3D CARDSTRUST THE
FIRST PULL.
The browser owns the deck order, chosen cards and upright or reversed orientation. Daily, Current Situation, Past–Present–Future and two-option Decision spreads provide one-, three- and five-card structures without changing who makes the selection.
VALIDATED CARDS.
A NEW READING.
The server accepts card, orientation and position identifiers, validates the spread and rebuilds canonical meanings from trusted project data before contacting OpenAI. The resulting overview, per-card interpretation, synthesis and reflection remain tied to the exact draw.
The revealed cards, reading title and opening interpretation stay visible together.
OPENAI / STRUCTURED RESULT
The complete reading ends with a synthesis, reflection prompt and optional generated narration.
OPENAI OR ELEVENLABS VOICETHE SAME
RITUAL, CLOSE.
Selection, revealed cards and detailed interpretation adapt to a narrow screen without removing the table, the deck artwork or the final reflection.
Choose the cards directly from the mobile fan.
Keep the spread and opening interpretation together.
Read the synthesis or play its generated narration.
ENTER THE READING
Veil & Arcana is live in the browser. The current MVP is English-only and limited to the Major Arcana. It is designed for reflection and entertainment, not professional advice.
OPEN TAROT — VEIL & ARCANA ↗