What it does
Accelerates
Drafts voice/multimodal interaction patterns consistent with the visual flow; flags where a modality switch changes the task's cognitive load
Agent 17 · 06 Experience Crafting
Multimodal Interaction Agent
Designs and evaluates voice, and other non-visual interaction modalities
What it never does
Stops here
Owns emotion-aware/biometric adaptation design decisions in sensitive contexts (health, wellness) without human clinical/ethical review
- Inputs
- Flow/IA structure, target modality (voice, gesture, context-aware switching)
- Outputs
- Voice-interaction scripts, modality-switching logic diagrams
- Model requirement
- Multimodal-capable model (voice + text reasoning)
Example prompt
"Draft a voice-interaction script for [task] that matches the visual flow's structure. Flag any step where voice changes the user's cognitive load versus the visual path."