feat(mobile): scaffold Expo mobile app for Timmy with Face/Matrix/Feed tabs
Task #42 — Timmy Harness: Expo Mobile App
## What was built
- New Expo artifact at artifacts/mobile, slug `mobile`, preview path `/mobile/`
- Three-tab bottom navigation (Face, Matrix, Feed) — NativeTabs with liquid glass on iOS 26+
- Dark wizard theme (#0A0A12 background, #7C3AED accent)
## WebSocket context (context/TimmyContext.tsx)
- Full WebSocket connection to /api/ws with exponential backoff reconnect (1s→30s cap)
- Sends visitor_enter handshake on connect, handles ping/pong
- Derives timmyMood from agent_state events (idle/thinking/working/speaking)
- recentEvents list capped at 100
- sendVisitorMessage() sets mood to "thinking" immediately on send (deterministic waiting state)
- speaking mood auto-reverts after estimated TTS duration
## Face tab (app/(tabs)/index.tsx)
- Animated 2D wizard face via react-native-svg (hat, head, beard, eyes, pupils, mouth arc, magic orb)
- AnimatedPupils: pupilScaleAnim drives actual rendered pupil Circle radius (BASE_PUPIL_R * scale)
- AnimatedEyelids: eyeScaleYAnim drives top eyelid overlay via Animated.Value listener
- AnimatedMouth: smileAnim + mouthOscAnim combined; SVG Path rebuilt on each frame via listener
- speaking mood: 1Hz mouth oscillation via Animated.loop; per-mood body bob speed/amplitude
- @react-native-voice/voice installed and statically imported; Voice.onSpeechResults wired properly
- startMicPulse/stopMicPulse declared before native voice useEffect (correct hook order)
- Web Speech API typed with SpeechRecognitionWindow local interface (zero `any` casts)
- sendVisitorMessage() called on final transcript (also triggers thinking mood immediately)
- expo-speech TTS speaks Timmy's chat replies on native
## Matrix tab (app/(tabs)/matrix.tsx)
- URL normalization: strips existing protocol, uses http for localhost, https for all other hosts
- Full-screen WebView with loading spinner and error/retry state; iframe fallback for web
## Feed tab (app/(tabs)/feed.tsx)
- FlatList<WsEvent> with proper generics; EventConfig discriminated union (Feather|MaterialCommunityIcons)
- Icon names typed via React.ComponentProps["name"] (no `any`)
- Color-coded events; event count in header; empty state with connection-aware message
## Type safety
- TypeScript typecheck passes with 0 errors
- No `any` casts anywhere in new code
## Deviations
- expo-av removed (not used; voice input handled via @react-native-voice/voice + Web Speech API)
- expo-speech/expo-av NOT in app.json plugins (no config plugins — causes PluginError if listed)
- app.json extra.apiDomain added for env-driven domain configuration per requirement
- expo-speech pinned ~14.0.8, react-native-webview 13.15.0 for Expo SDK 54 compat
- artifact.toml ensurePreviewReachable removed (Expo uses expo-domain router)
- @react-native-voice/voice works in Expo Go Android; iOS needs native build (graceful fallback)
Replit-Task-Id: 0748cbbf-7b84-4149-8fc0-9d697287a0e6
2026-03-19 23:55:16 +00:00
|
|
|
import {
|
|
|
|
|
Inter_400Regular,
|
|
|
|
|
Inter_500Medium,
|
|
|
|
|
Inter_600SemiBold,
|
|
|
|
|
Inter_700Bold,
|
|
|
|
|
useFonts,
|
|
|
|
|
} from "@expo-google-fonts/inter";
|
2026-03-23 14:31:47 +00:00
|
|
|
import AsyncStorage from "@react-native-async-storage/async-storage";
|
feat(mobile): scaffold Expo mobile app for Timmy with Face/Matrix/Feed tabs
Task #42 — Timmy Harness: Expo Mobile App
## What was built
- New Expo artifact at artifacts/mobile, slug `mobile`, preview path `/mobile/`
- Three-tab bottom navigation (Face, Matrix, Feed) — NativeTabs with liquid glass on iOS 26+
- Dark wizard theme (#0A0A12 background, #7C3AED accent)
## WebSocket context (context/TimmyContext.tsx)
- Full WebSocket connection to /api/ws with exponential backoff reconnect (1s→30s cap)
- Sends visitor_enter handshake on connect, handles ping/pong
- Derives timmyMood from agent_state events (idle/thinking/working/speaking)
- recentEvents list capped at 100
- sendVisitorMessage() sets mood to "thinking" immediately on send (deterministic waiting state)
- speaking mood auto-reverts after estimated TTS duration
## Face tab (app/(tabs)/index.tsx)
- Animated 2D wizard face via react-native-svg (hat, head, beard, eyes, pupils, mouth arc, magic orb)
- AnimatedPupils: pupilScaleAnim drives actual rendered pupil Circle radius (BASE_PUPIL_R * scale)
- AnimatedEyelids: eyeScaleYAnim drives top eyelid overlay via Animated.Value listener
- AnimatedMouth: smileAnim + mouthOscAnim combined; SVG Path rebuilt on each frame via listener
- speaking mood: 1Hz mouth oscillation via Animated.loop; per-mood body bob speed/amplitude
- @react-native-voice/voice installed and statically imported; Voice.onSpeechResults wired properly
- startMicPulse/stopMicPulse declared before native voice useEffect (correct hook order)
- Web Speech API typed with SpeechRecognitionWindow local interface (zero `any` casts)
- sendVisitorMessage() called on final transcript (also triggers thinking mood immediately)
- expo-speech TTS speaks Timmy's chat replies on native
## Matrix tab (app/(tabs)/matrix.tsx)
- URL normalization: strips existing protocol, uses http for localhost, https for all other hosts
- Full-screen WebView with loading spinner and error/retry state; iframe fallback for web
## Feed tab (app/(tabs)/feed.tsx)
- FlatList<WsEvent> with proper generics; EventConfig discriminated union (Feather|MaterialCommunityIcons)
- Icon names typed via React.ComponentProps["name"] (no `any`)
- Color-coded events; event count in header; empty state with connection-aware message
## Type safety
- TypeScript typecheck passes with 0 errors
- No `any` casts anywhere in new code
## Deviations
- expo-av removed (not used; voice input handled via @react-native-voice/voice + Web Speech API)
- expo-speech/expo-av NOT in app.json plugins (no config plugins — causes PluginError if listed)
- app.json extra.apiDomain added for env-driven domain configuration per requirement
- expo-speech pinned ~14.0.8, react-native-webview 13.15.0 for Expo SDK 54 compat
- artifact.toml ensurePreviewReachable removed (Expo uses expo-domain router)
- @react-native-voice/voice works in Expo Go Android; iOS needs native build (graceful fallback)
Replit-Task-Id: 0748cbbf-7b84-4149-8fc0-9d697287a0e6
2026-03-19 23:55:16 +00:00
|
|
|
import { QueryClient, QueryClientProvider } from "@tanstack/react-query";
|
2026-03-23 14:31:47 +00:00
|
|
|
import { Stack, router, useSegments } from "expo-router";
|
feat(mobile): scaffold Expo mobile app for Timmy with Face/Matrix/Feed tabs
Task #42 — Timmy Harness: Expo Mobile App
## What was built
- New Expo artifact at artifacts/mobile, slug `mobile`, preview path `/mobile/`
- Three-tab bottom navigation (Face, Matrix, Feed) — NativeTabs with liquid glass on iOS 26+
- Dark wizard theme (#0A0A12 background, #7C3AED accent)
## WebSocket context (context/TimmyContext.tsx)
- Full WebSocket connection to /api/ws with exponential backoff reconnect (1s→30s cap)
- Sends visitor_enter handshake on connect, handles ping/pong
- Derives timmyMood from agent_state events (idle/thinking/working/speaking)
- recentEvents list capped at 100
- sendVisitorMessage() sets mood to "thinking" immediately on send (deterministic waiting state)
- speaking mood auto-reverts after estimated TTS duration
## Face tab (app/(tabs)/index.tsx)
- Animated 2D wizard face via react-native-svg (hat, head, beard, eyes, pupils, mouth arc, magic orb)
- AnimatedPupils: pupilScaleAnim drives actual rendered pupil Circle radius (BASE_PUPIL_R * scale)
- AnimatedEyelids: eyeScaleYAnim drives top eyelid overlay via Animated.Value listener
- AnimatedMouth: smileAnim + mouthOscAnim combined; SVG Path rebuilt on each frame via listener
- speaking mood: 1Hz mouth oscillation via Animated.loop; per-mood body bob speed/amplitude
- @react-native-voice/voice installed and statically imported; Voice.onSpeechResults wired properly
- startMicPulse/stopMicPulse declared before native voice useEffect (correct hook order)
- Web Speech API typed with SpeechRecognitionWindow local interface (zero `any` casts)
- sendVisitorMessage() called on final transcript (also triggers thinking mood immediately)
- expo-speech TTS speaks Timmy's chat replies on native
## Matrix tab (app/(tabs)/matrix.tsx)
- URL normalization: strips existing protocol, uses http for localhost, https for all other hosts
- Full-screen WebView with loading spinner and error/retry state; iframe fallback for web
## Feed tab (app/(tabs)/feed.tsx)
- FlatList<WsEvent> with proper generics; EventConfig discriminated union (Feather|MaterialCommunityIcons)
- Icon names typed via React.ComponentProps["name"] (no `any`)
- Color-coded events; event count in header; empty state with connection-aware message
## Type safety
- TypeScript typecheck passes with 0 errors
- No `any` casts anywhere in new code
## Deviations
- expo-av removed (not used; voice input handled via @react-native-voice/voice + Web Speech API)
- expo-speech/expo-av NOT in app.json plugins (no config plugins — causes PluginError if listed)
- app.json extra.apiDomain added for env-driven domain configuration per requirement
- expo-speech pinned ~14.0.8, react-native-webview 13.15.0 for Expo SDK 54 compat
- artifact.toml ensurePreviewReachable removed (Expo uses expo-domain router)
- @react-native-voice/voice works in Expo Go Android; iOS needs native build (graceful fallback)
Replit-Task-Id: 0748cbbf-7b84-4149-8fc0-9d697287a0e6
2026-03-19 23:55:16 +00:00
|
|
|
import * as SplashScreen from "expo-splash-screen";
|
2026-03-23 14:31:47 +00:00
|
|
|
import React, { useEffect, useState } from "react";
|
feat(mobile): scaffold Expo mobile app for Timmy with Face/Matrix/Feed tabs
Task #42 — Timmy Harness: Expo Mobile App
## What was built
- New Expo artifact at artifacts/mobile, slug `mobile`, preview path `/mobile/`
- Three-tab bottom navigation (Face, Matrix, Feed) — NativeTabs with liquid glass on iOS 26+
- Dark wizard theme (#0A0A12 background, #7C3AED accent)
## WebSocket context (context/TimmyContext.tsx)
- Full WebSocket connection to /api/ws with exponential backoff reconnect (1s→30s cap)
- Sends visitor_enter handshake on connect, handles ping/pong
- Derives timmyMood from agent_state events (idle/thinking/working/speaking)
- recentEvents list capped at 100
- sendVisitorMessage() sets mood to "thinking" immediately on send (deterministic waiting state)
- speaking mood auto-reverts after estimated TTS duration
## Face tab (app/(tabs)/index.tsx)
- Animated 2D wizard face via react-native-svg (hat, head, beard, eyes, pupils, mouth arc, magic orb)
- AnimatedPupils: pupilScaleAnim drives actual rendered pupil Circle radius (BASE_PUPIL_R * scale)
- AnimatedEyelids: eyeScaleYAnim drives top eyelid overlay via Animated.Value listener
- AnimatedMouth: smileAnim + mouthOscAnim combined; SVG Path rebuilt on each frame via listener
- speaking mood: 1Hz mouth oscillation via Animated.loop; per-mood body bob speed/amplitude
- @react-native-voice/voice installed and statically imported; Voice.onSpeechResults wired properly
- startMicPulse/stopMicPulse declared before native voice useEffect (correct hook order)
- Web Speech API typed with SpeechRecognitionWindow local interface (zero `any` casts)
- sendVisitorMessage() called on final transcript (also triggers thinking mood immediately)
- expo-speech TTS speaks Timmy's chat replies on native
## Matrix tab (app/(tabs)/matrix.tsx)
- URL normalization: strips existing protocol, uses http for localhost, https for all other hosts
- Full-screen WebView with loading spinner and error/retry state; iframe fallback for web
## Feed tab (app/(tabs)/feed.tsx)
- FlatList<WsEvent> with proper generics; EventConfig discriminated union (Feather|MaterialCommunityIcons)
- Icon names typed via React.ComponentProps["name"] (no `any`)
- Color-coded events; event count in header; empty state with connection-aware message
## Type safety
- TypeScript typecheck passes with 0 errors
- No `any` casts anywhere in new code
## Deviations
- expo-av removed (not used; voice input handled via @react-native-voice/voice + Web Speech API)
- expo-speech/expo-av NOT in app.json plugins (no config plugins — causes PluginError if listed)
- app.json extra.apiDomain added for env-driven domain configuration per requirement
- expo-speech pinned ~14.0.8, react-native-webview 13.15.0 for Expo SDK 54 compat
- artifact.toml ensurePreviewReachable removed (Expo uses expo-domain router)
- @react-native-voice/voice works in Expo Go Android; iOS needs native build (graceful fallback)
Replit-Task-Id: 0748cbbf-7b84-4149-8fc0-9d697287a0e6
2026-03-19 23:55:16 +00:00
|
|
|
import { GestureHandlerRootView } from "react-native-gesture-handler";
|
|
|
|
|
import { KeyboardProvider } from "react-native-keyboard-controller";
|
|
|
|
|
import { SafeAreaProvider } from "react-native-safe-area-context";
|
|
|
|
|
|
|
|
|
|
import { ErrorBoundary } from "@/components/ErrorBoundary";
|
|
|
|
|
import { TimmyProvider } from "@/context/TimmyContext";
|
2026-03-23 14:31:47 +00:00
|
|
|
import { ONBOARDING_COMPLETED_KEY } from "@/constants/storage-keys";
|
feat(mobile): scaffold Expo mobile app for Timmy with Face/Matrix/Feed tabs
Task #42 — Timmy Harness: Expo Mobile App
## What was built
- New Expo artifact at artifacts/mobile, slug `mobile`, preview path `/mobile/`
- Three-tab bottom navigation (Face, Matrix, Feed) — NativeTabs with liquid glass on iOS 26+
- Dark wizard theme (#0A0A12 background, #7C3AED accent)
## WebSocket context (context/TimmyContext.tsx)
- Full WebSocket connection to /api/ws with exponential backoff reconnect (1s→30s cap)
- Sends visitor_enter handshake on connect, handles ping/pong
- Derives timmyMood from agent_state events (idle/thinking/working/speaking)
- recentEvents list capped at 100
- sendVisitorMessage() sets mood to "thinking" immediately on send (deterministic waiting state)
- speaking mood auto-reverts after estimated TTS duration
## Face tab (app/(tabs)/index.tsx)
- Animated 2D wizard face via react-native-svg (hat, head, beard, eyes, pupils, mouth arc, magic orb)
- AnimatedPupils: pupilScaleAnim drives actual rendered pupil Circle radius (BASE_PUPIL_R * scale)
- AnimatedEyelids: eyeScaleYAnim drives top eyelid overlay via Animated.Value listener
- AnimatedMouth: smileAnim + mouthOscAnim combined; SVG Path rebuilt on each frame via listener
- speaking mood: 1Hz mouth oscillation via Animated.loop; per-mood body bob speed/amplitude
- @react-native-voice/voice installed and statically imported; Voice.onSpeechResults wired properly
- startMicPulse/stopMicPulse declared before native voice useEffect (correct hook order)
- Web Speech API typed with SpeechRecognitionWindow local interface (zero `any` casts)
- sendVisitorMessage() called on final transcript (also triggers thinking mood immediately)
- expo-speech TTS speaks Timmy's chat replies on native
## Matrix tab (app/(tabs)/matrix.tsx)
- URL normalization: strips existing protocol, uses http for localhost, https for all other hosts
- Full-screen WebView with loading spinner and error/retry state; iframe fallback for web
## Feed tab (app/(tabs)/feed.tsx)
- FlatList<WsEvent> with proper generics; EventConfig discriminated union (Feather|MaterialCommunityIcons)
- Icon names typed via React.ComponentProps["name"] (no `any`)
- Color-coded events; event count in header; empty state with connection-aware message
## Type safety
- TypeScript typecheck passes with 0 errors
- No `any` casts anywhere in new code
## Deviations
- expo-av removed (not used; voice input handled via @react-native-voice/voice + Web Speech API)
- expo-speech/expo-av NOT in app.json plugins (no config plugins — causes PluginError if listed)
- app.json extra.apiDomain added for env-driven domain configuration per requirement
- expo-speech pinned ~14.0.8, react-native-webview 13.15.0 for Expo SDK 54 compat
- artifact.toml ensurePreviewReachable removed (Expo uses expo-domain router)
- @react-native-voice/voice works in Expo Go Android; iOS needs native build (graceful fallback)
Replit-Task-Id: 0748cbbf-7b84-4149-8fc0-9d697287a0e6
2026-03-19 23:55:16 +00:00
|
|
|
|
|
|
|
|
SplashScreen.preventAutoHideAsync();
|
|
|
|
|
|
|
|
|
|
const queryClient = new QueryClient();
|
|
|
|
|
|
|
|
|
|
function RootLayoutNav() {
|
2026-03-23 14:31:47 +00:00
|
|
|
const segments = useSegments();
|
|
|
|
|
const [onboardingChecked, setOnboardingChecked] = useState(false);
|
|
|
|
|
const [needsOnboarding, setNeedsOnboarding] = useState(false);
|
|
|
|
|
|
|
|
|
|
useEffect(() => {
|
|
|
|
|
AsyncStorage.getItem(ONBOARDING_COMPLETED_KEY).then((value) => {
|
|
|
|
|
setNeedsOnboarding(value !== "true");
|
|
|
|
|
setOnboardingChecked(true);
|
|
|
|
|
});
|
|
|
|
|
}, []);
|
|
|
|
|
|
|
|
|
|
useEffect(() => {
|
|
|
|
|
if (!onboardingChecked) return;
|
|
|
|
|
|
|
|
|
|
const inOnboarding = segments[0] === "onboarding";
|
|
|
|
|
|
|
|
|
|
if (needsOnboarding && !inOnboarding) {
|
|
|
|
|
router.replace("/onboarding");
|
|
|
|
|
} else if (!needsOnboarding && inOnboarding) {
|
|
|
|
|
router.replace("/(tabs)");
|
|
|
|
|
}
|
|
|
|
|
}, [onboardingChecked, needsOnboarding, segments]);
|
|
|
|
|
|
feat(mobile): scaffold Expo mobile app for Timmy with Face/Matrix/Feed tabs
Task #42 — Timmy Harness: Expo Mobile App
## What was built
- New Expo artifact at artifacts/mobile, slug `mobile`, preview path `/mobile/`
- Three-tab bottom navigation (Face, Matrix, Feed) — NativeTabs with liquid glass on iOS 26+
- Dark wizard theme (#0A0A12 background, #7C3AED accent)
## WebSocket context (context/TimmyContext.tsx)
- Full WebSocket connection to /api/ws with exponential backoff reconnect (1s→30s cap)
- Sends visitor_enter handshake on connect, handles ping/pong
- Derives timmyMood from agent_state events (idle/thinking/working/speaking)
- recentEvents list capped at 100
- sendVisitorMessage() sets mood to "thinking" immediately on send (deterministic waiting state)
- speaking mood auto-reverts after estimated TTS duration
## Face tab (app/(tabs)/index.tsx)
- Animated 2D wizard face via react-native-svg (hat, head, beard, eyes, pupils, mouth arc, magic orb)
- AnimatedPupils: pupilScaleAnim drives actual rendered pupil Circle radius (BASE_PUPIL_R * scale)
- AnimatedEyelids: eyeScaleYAnim drives top eyelid overlay via Animated.Value listener
- AnimatedMouth: smileAnim + mouthOscAnim combined; SVG Path rebuilt on each frame via listener
- speaking mood: 1Hz mouth oscillation via Animated.loop; per-mood body bob speed/amplitude
- @react-native-voice/voice installed and statically imported; Voice.onSpeechResults wired properly
- startMicPulse/stopMicPulse declared before native voice useEffect (correct hook order)
- Web Speech API typed with SpeechRecognitionWindow local interface (zero `any` casts)
- sendVisitorMessage() called on final transcript (also triggers thinking mood immediately)
- expo-speech TTS speaks Timmy's chat replies on native
## Matrix tab (app/(tabs)/matrix.tsx)
- URL normalization: strips existing protocol, uses http for localhost, https for all other hosts
- Full-screen WebView with loading spinner and error/retry state; iframe fallback for web
## Feed tab (app/(tabs)/feed.tsx)
- FlatList<WsEvent> with proper generics; EventConfig discriminated union (Feather|MaterialCommunityIcons)
- Icon names typed via React.ComponentProps["name"] (no `any`)
- Color-coded events; event count in header; empty state with connection-aware message
## Type safety
- TypeScript typecheck passes with 0 errors
- No `any` casts anywhere in new code
## Deviations
- expo-av removed (not used; voice input handled via @react-native-voice/voice + Web Speech API)
- expo-speech/expo-av NOT in app.json plugins (no config plugins — causes PluginError if listed)
- app.json extra.apiDomain added for env-driven domain configuration per requirement
- expo-speech pinned ~14.0.8, react-native-webview 13.15.0 for Expo SDK 54 compat
- artifact.toml ensurePreviewReachable removed (Expo uses expo-domain router)
- @react-native-voice/voice works in Expo Go Android; iOS needs native build (graceful fallback)
Replit-Task-Id: 0748cbbf-7b84-4149-8fc0-9d697287a0e6
2026-03-19 23:55:16 +00:00
|
|
|
return (
|
|
|
|
|
<Stack screenOptions={{ headerBackTitle: "Back" }}>
|
2026-03-23 14:31:47 +00:00
|
|
|
<Stack.Screen name="onboarding" options={{ headerShown: false, animation: "none" }} />
|
feat(mobile): scaffold Expo mobile app for Timmy with Face/Matrix/Feed tabs
Task #42 — Timmy Harness: Expo Mobile App
## What was built
- New Expo artifact at artifacts/mobile, slug `mobile`, preview path `/mobile/`
- Three-tab bottom navigation (Face, Matrix, Feed) — NativeTabs with liquid glass on iOS 26+
- Dark wizard theme (#0A0A12 background, #7C3AED accent)
## WebSocket context (context/TimmyContext.tsx)
- Full WebSocket connection to /api/ws with exponential backoff reconnect (1s→30s cap)
- Sends visitor_enter handshake on connect, handles ping/pong
- Derives timmyMood from agent_state events (idle/thinking/working/speaking)
- recentEvents list capped at 100
- sendVisitorMessage() sets mood to "thinking" immediately on send (deterministic waiting state)
- speaking mood auto-reverts after estimated TTS duration
## Face tab (app/(tabs)/index.tsx)
- Animated 2D wizard face via react-native-svg (hat, head, beard, eyes, pupils, mouth arc, magic orb)
- AnimatedPupils: pupilScaleAnim drives actual rendered pupil Circle radius (BASE_PUPIL_R * scale)
- AnimatedEyelids: eyeScaleYAnim drives top eyelid overlay via Animated.Value listener
- AnimatedMouth: smileAnim + mouthOscAnim combined; SVG Path rebuilt on each frame via listener
- speaking mood: 1Hz mouth oscillation via Animated.loop; per-mood body bob speed/amplitude
- @react-native-voice/voice installed and statically imported; Voice.onSpeechResults wired properly
- startMicPulse/stopMicPulse declared before native voice useEffect (correct hook order)
- Web Speech API typed with SpeechRecognitionWindow local interface (zero `any` casts)
- sendVisitorMessage() called on final transcript (also triggers thinking mood immediately)
- expo-speech TTS speaks Timmy's chat replies on native
## Matrix tab (app/(tabs)/matrix.tsx)
- URL normalization: strips existing protocol, uses http for localhost, https for all other hosts
- Full-screen WebView with loading spinner and error/retry state; iframe fallback for web
## Feed tab (app/(tabs)/feed.tsx)
- FlatList<WsEvent> with proper generics; EventConfig discriminated union (Feather|MaterialCommunityIcons)
- Icon names typed via React.ComponentProps["name"] (no `any`)
- Color-coded events; event count in header; empty state with connection-aware message
## Type safety
- TypeScript typecheck passes with 0 errors
- No `any` casts anywhere in new code
## Deviations
- expo-av removed (not used; voice input handled via @react-native-voice/voice + Web Speech API)
- expo-speech/expo-av NOT in app.json plugins (no config plugins — causes PluginError if listed)
- app.json extra.apiDomain added for env-driven domain configuration per requirement
- expo-speech pinned ~14.0.8, react-native-webview 13.15.0 for Expo SDK 54 compat
- artifact.toml ensurePreviewReachable removed (Expo uses expo-domain router)
- @react-native-voice/voice works in Expo Go Android; iOS needs native build (graceful fallback)
Replit-Task-Id: 0748cbbf-7b84-4149-8fc0-9d697287a0e6
2026-03-19 23:55:16 +00:00
|
|
|
<Stack.Screen name="(tabs)" options={{ headerShown: false }} />
|
2026-03-23 22:07:04 +00:00
|
|
|
<Stack.Screen name="settings" options={{ headerShown: false, presentation: "modal" }} />
|
feat(mobile): scaffold Expo mobile app for Timmy with Face/Matrix/Feed tabs
Task #42 — Timmy Harness: Expo Mobile App
## What was built
- New Expo artifact at artifacts/mobile, slug `mobile`, preview path `/mobile/`
- Three-tab bottom navigation (Face, Matrix, Feed) — NativeTabs with liquid glass on iOS 26+
- Dark wizard theme (#0A0A12 background, #7C3AED accent)
## WebSocket context (context/TimmyContext.tsx)
- Full WebSocket connection to /api/ws with exponential backoff reconnect (1s→30s cap)
- Sends visitor_enter handshake on connect, handles ping/pong
- Derives timmyMood from agent_state events (idle/thinking/working/speaking)
- recentEvents list capped at 100
- sendVisitorMessage() sets mood to "thinking" immediately on send (deterministic waiting state)
- speaking mood auto-reverts after estimated TTS duration
## Face tab (app/(tabs)/index.tsx)
- Animated 2D wizard face via react-native-svg (hat, head, beard, eyes, pupils, mouth arc, magic orb)
- AnimatedPupils: pupilScaleAnim drives actual rendered pupil Circle radius (BASE_PUPIL_R * scale)
- AnimatedEyelids: eyeScaleYAnim drives top eyelid overlay via Animated.Value listener
- AnimatedMouth: smileAnim + mouthOscAnim combined; SVG Path rebuilt on each frame via listener
- speaking mood: 1Hz mouth oscillation via Animated.loop; per-mood body bob speed/amplitude
- @react-native-voice/voice installed and statically imported; Voice.onSpeechResults wired properly
- startMicPulse/stopMicPulse declared before native voice useEffect (correct hook order)
- Web Speech API typed with SpeechRecognitionWindow local interface (zero `any` casts)
- sendVisitorMessage() called on final transcript (also triggers thinking mood immediately)
- expo-speech TTS speaks Timmy's chat replies on native
## Matrix tab (app/(tabs)/matrix.tsx)
- URL normalization: strips existing protocol, uses http for localhost, https for all other hosts
- Full-screen WebView with loading spinner and error/retry state; iframe fallback for web
## Feed tab (app/(tabs)/feed.tsx)
- FlatList<WsEvent> with proper generics; EventConfig discriminated union (Feather|MaterialCommunityIcons)
- Icon names typed via React.ComponentProps["name"] (no `any`)
- Color-coded events; event count in header; empty state with connection-aware message
## Type safety
- TypeScript typecheck passes with 0 errors
- No `any` casts anywhere in new code
## Deviations
- expo-av removed (not used; voice input handled via @react-native-voice/voice + Web Speech API)
- expo-speech/expo-av NOT in app.json plugins (no config plugins — causes PluginError if listed)
- app.json extra.apiDomain added for env-driven domain configuration per requirement
- expo-speech pinned ~14.0.8, react-native-webview 13.15.0 for Expo SDK 54 compat
- artifact.toml ensurePreviewReachable removed (Expo uses expo-domain router)
- @react-native-voice/voice works in Expo Go Android; iOS needs native build (graceful fallback)
Replit-Task-Id: 0748cbbf-7b84-4149-8fc0-9d697287a0e6
2026-03-19 23:55:16 +00:00
|
|
|
</Stack>
|
|
|
|
|
);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
export default function RootLayout() {
|
|
|
|
|
const [fontsLoaded, fontError] = useFonts({
|
|
|
|
|
Inter_400Regular,
|
|
|
|
|
Inter_500Medium,
|
|
|
|
|
Inter_600SemiBold,
|
|
|
|
|
Inter_700Bold,
|
|
|
|
|
});
|
|
|
|
|
|
|
|
|
|
useEffect(() => {
|
|
|
|
|
if (fontsLoaded || fontError) {
|
|
|
|
|
SplashScreen.hideAsync();
|
|
|
|
|
}
|
|
|
|
|
}, [fontsLoaded, fontError]);
|
|
|
|
|
|
|
|
|
|
if (!fontsLoaded && !fontError) return null;
|
|
|
|
|
|
|
|
|
|
return (
|
|
|
|
|
<SafeAreaProvider>
|
|
|
|
|
<ErrorBoundary>
|
|
|
|
|
<QueryClientProvider client={queryClient}>
|
|
|
|
|
<GestureHandlerRootView style={{ flex: 1 }}>
|
|
|
|
|
<KeyboardProvider>
|
|
|
|
|
<TimmyProvider>
|
|
|
|
|
<RootLayoutNav />
|
|
|
|
|
</TimmyProvider>
|
|
|
|
|
</KeyboardProvider>
|
|
|
|
|
</GestureHandlerRootView>
|
|
|
|
|
</QueryClientProvider>
|
|
|
|
|
</ErrorBoundary>
|
|
|
|
|
</SafeAreaProvider>
|
|
|
|
|
);
|
|
|
|
|
}
|