Voice & intents

Voice & multimodal intents

Visual tokens don't apply to Siri, Shortcuts, Alexa, or CarPlay. Intent schemas + utterance copy are the design system here.

Rules

  1. Same entity names as ENTITY_NAMES.md
  2. Same error shape as COPY_STANDARDS.md (spoken: short, actionable)
  3. No PII in spoken confirmations ("Meal logged" not "Logged pasta at 7:42 PM with location")

Tasteful iOS (implemented)

IntentUser phrase (examples)Spoken / compact result
Meal suggestion"What should I eat?"Opens in-app recommendation — avoid stacking dialogs
Log meal"Log a meal"Opens unified add sheet
Favorites"Show favorite meals"Navigates to Meals (favorites)

Reference: PRODUCT_BRIEF.md → Siri & App Shortcuts

Future surfaces

SurfaceDesign artifact
watchOSSubset of foundation spacing; large tap targets; glances not forms
tvOSFocus-engine lists; 10-foot type scale multiplier
CarPlayTemplate lists only — pattern doc, no custom chrome
Alexa / GoogleSkill manifest + utterance table in this folder per app

Template: new voice intent

### Intent: [Name]
- **Entities:** (from ENTITY_NAMES.md)
- **Utterances:** …
- **Success:** …
- **Failure:** …
- **Opens app?** yes/no/deep link

Add a row here before shipping any new shortcut or voice skill.