Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
24 commits
Select commit Hold shift + click to select a range
0924e0a
feat: add Watch <-> iPhone voice command communication
RoyGuanyu Feb 12, 2026
b9785d6
feat: add press-to-talk voice input, OpenClaw client, and device regi…
RoyGuanyu Feb 12, 2026
951720d
feat: wire voice→genUI full pipeline, hands-free mode, Watch bridge
RoyGuanyu Feb 13, 2026
d2ea43d
feat: add ClawFree demo landing page
RoyGuanyu Feb 13, 2026
bca4985
feat: dark theme, brand colors, UI polish for iPhone/iPad/macOS layouts
RoyGuanyu Feb 13, 2026
e10eda2
修復 macOS build:升級 record ^6.0.0 + 加回 WatchCompanion target
RoyGuanyu Feb 13, 2026
ba55a69
feat: UI/UX polish — glassmorphism, animations, typography upgrade
RoyGuanyu Feb 13, 2026
26b695c
feat: DeviceRegistry mock fallback + 裝置管理 UI 頁面
RoyGuanyu Feb 13, 2026
db5e370
feat: Watch TTS 語音回覆 + 對話式 UI 改進
RoyGuanyu Feb 13, 2026
13a6959
feat: App Icon dark/light/tinted 三組版本
RoyGuanyu Feb 13, 2026
3b36dfc
fix: genUI v0.9 API rename + track migration branch
RoyGuanyu Feb 13, 2026
8bc4491
fix: TTS race condition with generation counter
RoyGuanyu Feb 13, 2026
9ece068
fix: Watch build — target dependency, embed phase, Info.plist key
RoyGuanyu Feb 13, 2026
b37265d
feat: Watch gateway relay + multi-device communication
RoyGuanyu Feb 13, 2026
955cad8
fix: Watch UX — inline input, app icon, message relay buffering
RoyGuanyu Feb 13, 2026
c4fd84b
feat: Voice-first Watch UX + InputCoordinator + Watch management
RoyGuanyu Feb 13, 2026
2bc7001
feat(watch): one-tap voice UX — big mic button with dictation auto-send
RoyGuanyu Feb 13, 2026
8abca85
feat: local WebSocket sync + trip planning demo responses
RoyGuanyu Feb 13, 2026
b5cf132
test: add Watch bridge voice_command + demo mode trip scenario tests
RoyGuanyu Feb 14, 2026
3f1ade7
feat: auto-demo mode + deep link input for demo recording
RoyGuanyu Feb 14, 2026
abe3f91
feat(watch): 互動式 UI — 建 Agent 配置流程 + 旅行規劃流程
RoyGuanyu Feb 14, 2026
12fadbc
feat: Watch UI 即時同步到 iPhone/iPad/macOS
RoyGuanyu Feb 14, 2026
8835cef
feat: English UI + Watch voice guide + sync tests
RoyGuanyu Feb 14, 2026
2f9c7be
feat: Watch auto-demo mode + XCUITest scaffold
RoyGuanyu Feb 15, 2026
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
141 changes: 141 additions & 0 deletions docs/DEMO_SCRIPT.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,141 @@
# ClawFree — 3-Minute Demo Video Script

## Overview
- **Duration**: 3:00
- **Target**: Anthropic "Built with Opus 4.6" Hackathon judges
- **Scoring**: Demo 30% | Impact 25% | Opus 4.6 Use 25% | Depth 20%
- **Devices**: MacBook (macOS), iPhone, iPad, Apple Watch

---

## Scene 1: Opening Hook (0:00 - 0:15)

**Visual**: Dark screen → ClawFree logo fade in → tagline appears

**Voiceover**:
> "What if your AI wasn't trapped behind a keyboard? What if you could talk to it from your watch, your phone, your tablet — and it built the UI for you in real-time?"

**Text on screen**: "ClawFree — Hands-free AI, everywhere you are"

---

## Scene 2: Architecture Overview (0:15 - 0:30)

**Visual**: Animated diagram showing 4 devices → OpenClaw Gateway → Opus 4.6 → genUI Surface

**Voiceover**:
> "ClawFree connects your Apple Watch, iPhone, iPad, and Mac to OpenClaw — your local AI gateway. Every interaction is powered by Opus 4.6, which generates rich, interactive UIs on the fly using our A2UI protocol."

---

## Scene 3: Voice Trip Planning — Watch (0:30 - 1:15)

**Visual**: Apple Watch on wrist / Watch simulator

**Action**:
1. Tap mic on Watch → speak: "Plan a 3-day trip to Tokyo"
2. Watch shows "Sending..." → "Planning..."
3. Cut to iPhone showing genUI Surface rendering:
- Day-by-day itinerary cards
- Budget breakdown
- Map preview
4. TTS plays: "Here's a 3-day Tokyo itinerary..."

**Voiceover**:
> "Start from your watch. Just speak — Opus 4.6 generates a complete travel planner with interactive cards, budget estimates, and daily schedules. No templates. Every UI is generated fresh."

**Key moment**: Show the A2UI JSON briefly → transformed into beautiful UI

---

## Scene 4: Voice Follow-up — iPhone (1:15 - 1:45)

**Visual**: iPhone screen

**Action**:
1. VoiceOrb pulses → user speaks: "Add Osaka on day 2"
2. genUI Surface updates — Osaka card appears in Day 2
3. User speaks: "What's the budget now?"
4. Budget card updates with new total
5. User taps "Book this itinerary" button on genUI surface

**Voiceover**:
> "Continue the conversation hands-free. Ask follow-ups, and the UI evolves. Mix voice and touch — tap to confirm, speak to explore. Opus 4.6 maintains context across turns."

---

## Scene 5: Control Tower — macOS (1:45 - 2:15)

**Visual**: macOS dashboard (TabletLayout)

**Action**:
1. Show full Control Tower: sidebar with agents, telemetry header with health vitals
2. Connected devices showing: iPhone ✅, Watch ✅, iPad ✅
3. Main surface showing the trip plan from iPhone
4. Type or speak: "Create an agent called TripBot specialized in Japan travel"
5. genUI renders agent configuration form
6. Fill settings → Save

**Voiceover**:
> "The Mac is your command center. See every connected device, monitor your AI's health, and manage agents. Here we're creating a specialized travel agent — Opus 4.6 generates the entire configuration UI from a single voice command."

---

## Scene 6: Multi-device Sync — iPad (2:15 - 2:35)

**Visual**: iPad simulator showing TabletLayout

**Action**:
1. iPad shows same trip plan synced from iPhone session
2. Pinch to zoom on map
3. Voice: "Show me hotel options near Shibuya"
4. genUI renders hotel comparison cards

**Voiceover**:
> "Every device stays in sync through OpenClaw. Pick up where you left off on any screen. The iPad's larger display shows the full power of genUI surfaces."

---

## Scene 7: The Vision — Ambient Computing (2:35 - 2:50)

**Visual**: Split screen showing all 4 devices simultaneously

**Voiceover**:
> "This is ambient computing. Your AI is always there — on your wrist for quick commands, in your pocket for deeper interactions, on your desk for full control. No cloud dependency — everything runs on your local network. Today it's powered by Opus 4.6. Tomorrow, it could run on local models for zero-cost, always-on intelligence."

---

## Scene 8: Closing (2:50 - 3:00)

**Visual**: ClawFree logo + tech stack badges + GitHub link

**Text on screen**:
- "Built with Opus 4.6 + genUI v0.9 + OpenClaw"
- "Team genUIne: Mike & Roy"
- GitHub: github.com/dart-technologies/clawfree

**Voiceover**:
> "ClawFree. Your AI, unbound."

---

## Production Notes

### Recording Tips
- Screen record each device separately, composite in editor
- Use QuickTime for Mac/iPad simulator recording
- iPhone: either real device or simulator
- Watch: simulator is fine (hardware not required)
- Add subtle background music (lo-fi / ambient)
- Transitions: cross-dissolve between scenes

### Demo Mode
- Use `DEMO_MODE=true` for cached responses (reliable)
- For Scene 3-4, consider one LIVE Opus 4.6 call to show real generation
- Have backup cached responses ready

### Tools
- **Screen recording**: QuickTime Player / OBS
- **Video editing**: iMovie / DaVinci Resolve (free)
- **Captions**: Add key text overlays for architecture/protocol
- **Music**: freemusicarchive.org or YouTube Audio Library
67 changes: 67 additions & 0 deletions docs/PITCH_SPEECH.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,67 @@
# ClawFree — Demo Day Pitch Speech

> Duration: ~2 minutes (before or after 3-min video)
> Speaker: Roy (with Mike)

---

## The Speech

Hi everyone, I'm Roy, and this is Mike. We're **Team genUIne**.

We built **ClawFree** — a hands-free, voice-powered AI interface that runs on your watch, your phone, your tablet, and your Mac. All powered by **Opus 4.6**.

But first, let me tell you how we got here.

### How We Work

Mike and I sync for **15 minutes a day**. That's it. The rest of the time, we iterate with AI. We don't waste time on meetings. We deliver, review, iterate.

Mike brought the vision — what if AI could generate its own UI? He built **genUI**, an A2UI protocol that lets Opus 4.6 create rich, interactive interfaces from scratch. No templates. No pre-built screens.

I brought the voice — what if you never had to touch a keyboard? I built the multi-device voice layer with OpenClaw, so you can speak from your Watch and see the result on your iPad.

### Why Opus 4.6 — Everywhere

Here's what makes us different. Most teams use Opus for one thing — a chatbot, a tool, a feature.

**We used Opus 4.6 for everything.**

- **Ideation**: We brainstormed with Opus. Evaluated impact. Refined our pitch.
- **Development**: Opus wrote the code. We iterated with voice commands — literally zero keyboard.
- **Product**: Opus generates every UI at runtime through genUI. Every screen you'll see was created by Opus 4.6 in real-time.
- **Marketing**: This website, this demo script, even this speech — Opus helped craft all of it.
- **Validation**: We used Opus to work backwards from judging criteria, identify gaps, and strengthen our weak spots.

The human decides **what matters**. Opus 4.6 figures out **how to make it happen**.

### The Vision

In the AI era, the question isn't "how do we build it?" — AI can figure that out. The question is: **"What should we build that influences the most people?"**

ClawFree is our answer. Ambient AI. On your wrist, in your pocket, on your desk. Always there, always ready. No keyboard required.

Today it runs on Opus 4.6. Tomorrow, with local models on a Mac Studio, it becomes always-on, zero-cost intelligence.

### Closing

We're excited to be here. Thank you to Anthropic, to Claude, to the Cerebral Valley community.

In this AI era, there are more challenges than ever — but also more opportunities than ever. And we're thrilled to be building at this frontier.

**ClawFree. Your AI, unbound.**

Thank you.

---

## Speaker Notes

- **Tone**: Confident but humble. Excited but grounded.
- **Pace**: Don't rush. Let key lines land.
- **Key moments to emphasize**:
- "15 minutes a day" — pause, let it sink in
- "Zero keyboard" — demonstrate with gesture
- "What should we build that influences the most people?" — this is the thesis
- **If time is short**: Cut the "How We Work" section, jump straight to Opus 4.6 usage
- **Language**: Primarily English for judges. Roy can mix in passion/energy naturally.
Loading