Add personal AI companion component with voice and visual states

[Copilot]

Implements a cross-UI AI companion with contextual visual states, speech synthesis, and wave-form visualization. Component appears as a pulsating orb with animated robotic face that responds to interaction states.

Implementation

Core Component (heady_companion.js)

• Vanilla JS class, zero dependencies (~20KB)
• Four visual states: idle, listening, speaking, thinking (CSS animations)
• Web Speech API integration with voice selection
• Canvas-based wave-form mouth animation during speech
• Configurable positioning, sizing, theming

Integration

• Drop-in component via script include or auto-init
• Integrated into heady_gaia_monitor.html as reference implementation
• Standalone demo page with customization controls

Customization

• System voice selection from available voices
• Color theming (primary/secondary/accent)
• Positioning (4 corners) and sizing (S/M/L)
• Placeholder for custom voice training

Usage

const companion = new HeadyCompanion({
    container: '#app',
    position: 'bottom-right',
    primaryColor: '#97c97c',
    voice: 'Google US English'
});

companion.speak('Hello! How can I help you today?');

Screenshots

Visual States

Idle	Listening	Speaking	Thinking

Integrated with HeadyGaia Monitor

Technical Notes

• Speech synthesis runs entirely in-browser (no external APIs)
• Wave-form visualization uses requestAnimationFrame for 60fps
• Proper cleanup in destroy() prevents memory leaks
• CodeQL: 0 alerts

Original prompt

Pull request: #24
nice. now across all user UI and ux, the personal ai companion should function across perceived boundaries and operate alongside the person present in their user space. I imagine a welcoming roundish symmetrical and pulsating relative to functioning certain ways such as when speaking the visual representation of the ai companion speaking is a mini friendly looking robotic style face that has an area where the mouth should be that shows moving wave type functions to represent the sound eminating from the device. this ai interface should be customizable to allow for ease in comfort including having many voice options including trading a model to have my voice for fun

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

[Copilot]

Copilot wasn't able to review any files in this pull request.