A menu's job no longer ends at the table. The dinner menu below is doing three jobs at once: persuading the guest holding it, surfacing when someone nearby searches for exactly what you serve, and answering correctly when an AI assistant is asked where to eat. One source drives all three — so what the guest reads and what the machines repeat can never drift apart. And it's not a mockup: view source on this page and the proof is right there.
Flip the switch // see what machines see
One artifact // two readers
Specimen № 01
Dinner
Set for guests · set for machines
First Movements
// schema.org/MenuSection
Bread & Cultured Butter9
sourdough, smoked salt
MenuItem.name · offers.price
Chicories & Anchovy14
winter greens, garlic crumb
MenuItem.name · offers.price
Crispy Oyster Mushrooms13
chili glaze, lime
MenuItem.name · offers.price
The Main Room
// schema.org/MenuSection
Half Roast Chicken29
pan jus, charred lemon
MenuItem.name · offers.price
Market Fish34
brown butter, capers
MenuItem.name · offers.price
Braised Short Rib38
potato purée, gremolata
MenuItem.name · offers.price
Squash Agnolotti26
sage, brown butter, hazelnut
MenuItem.name · offers.price
Last Light
// schema.org/MenuSection
Burnt-Honey Panna Cotta11
MenuItem.name · offers.price
Chocolate Tart12
olive oil, flake salt
MenuItem.name · offers.price
The proof // view source
Every line above is emitted a second time as structured, machine-readable data—one single source rendering both layers. The menu the guest reads and the menu the AI reads can never disagree. View source on this page and the proof is right there.