Lumiere

A cross-platform AI chat client built with React Native (Expo SDK 54) and TypeScript. Connects to 12+ AI providers including OpenClaw, Claude, OpenRouter, OpenAI, Gemini, Ollama, Kimi, Apple Intelligence, and more through a unified provider abstraction. Runs on iOS, Android, Web, Desktop (Electron), and Chrome Extension.

Website • Features • AI Providers • Screens • Screenshots • Getting Started • Tech Stack • Architecture • Testing • CI/CD • Internationalization • Contributing • License

Features

Core

Multi-provider support — Connect to 12+ AI providers including OpenClaw, Claude, OpenRouter, OpenAI, Gemini, Ollama, Apple Intelligence, and more
Multi-server management — Configure, switch between, and backup/restore multiple AI servers
Real-time streaming — Streamed AI responses with markdown rendering
Image attachments — Attach images, documents, and other files to chat messages (provider-dependent)
Message queue — Queue messages while the agent is responding
Slash commands — 38 built-in commands with autocomplete
Missions (Beta) — Multi-step AI-driven task planning with subtask execution and streaming progress
Skills — Teach agents custom capabilities via skill packs (OpenClaw)

Voice & Input

Voice transcription — Dictate messages using native iOS speech recognition
Recording overlay — Visual feedback with real-time transcription preview
Text-to-speech — Play/pause TTS on any message

Personalization

Color themes — 8 color palettes (Default, Pink, Green, Red, Blue, Purple, Orange, Glass)
Light & dark modes — System-aware theming with manual override
Session aliases — Custom display names for chat sessions
11 languages — English, Spanish, French, German, Hindi, Japanese, Korean, Portuguese (BR), Russian, Simplified Chinese, Traditional Chinese

Automation

Triggers — Create deep links that auto-send messages to specific sessions
Quick actions — Home screen shortcuts (iOS 3D Touch / Android) linked to triggers
Cron scheduler — Schedule recurring agent tasks (OpenClaw only)
Background notifications — Get notified of new messages when app is backgrounded

Security

Face ID / Touch ID — Biometric authentication on app launch and resume
Secure storage — API keys and tokens stored in the device keychain
Encrypted web storage — AES-GCM encryption via Web Crypto API for server credentials on web

Platform

iOS & Android — Full mobile support
iPad — Responsive layouts with form sheet modals on tablet
Web — Browser support via Metro bundler
Desktop — Electron app builds
Chrome Extension — Browser extension build
Deep linking — Open screens and execute triggers via lumiere:// URLs

AI Providers

Provider	Type	Chat	Images	Sessions	History	Scheduler	Skills
OpenClaw	`molt`	✅	✅	✅	✅	✅	✅
Claude	`claude`	✅	✅	—	—	—	—
OpenRouter	`openrouter`	✅	✅	—	—	—	—
OpenAI	`openai`	✅	✅	—	—	—	—
OpenAI Compatible	`openai-compatible`	✅	✅	—	—	—	—
Gemini	`gemini`	✅	✅	—	—	—	—
Gemini Nano	`gemini-nano`	✅	—	—	—	—	—
Kimi	`kimi`	✅	✅	—	—	—	—
Ollama	`ollama`	✅	—	—	—	—	—
Apple Intelligence	`apple`	✅	—	—	—	—	—
Echo Server	`echo`	✅	—	—	—	—	—

OpenClaw (Molt Gateway)

Full-featured provider with WebSocket streaming, server-side sessions, chat history persistence, cron job scheduling, and skill support. Requires a running Molt Gateway instance.

Claude (Anthropic API)

Direct integration with Anthropic's Messages API. Supports streaming responses and image attachments. Requires an Anthropic API key.

OpenRouter

Access hundreds of models from multiple providers through a single API. Supports streaming and image attachments. Requires an OpenRouter API key.

OpenAI

Direct integration with OpenAI's Chat Completions API. Supports streaming and image attachments. Requires an OpenAI API key.

OpenAI Compatible

Connect to any provider that implements the OpenAI-compatible API format (e.g., LM Studio, vLLM, Together AI).

Gemini

Google's Gemini models with streaming and image support. Requires a Google AI API key.

Gemini Nano

On-device AI using Google's Gemini Nano model. Android only. No external server needed.

Kimi

Moonshot AI's Kimi models with streaming and image support. Requires a Kimi API key.

Ollama

Connect to a local Ollama instance for offline AI. Includes a model selection screen to switch between installed models.

Apple Intelligence

On-device AI using Apple Foundation Models. Requires iOS 18+ with Apple Intelligence support. No external server needed.

Echo Server

Testing provider that echoes messages back. Useful for development and debugging.

Screens

Screen	File	Description
Chat	`app/index.tsx`	Main chat interface with streaming responses, markdown rendering, voice input, slash commands, and message favoriting
Settings	`app/settings.tsx`	App configuration including appearance, security, notifications, and account management
Servers	`app/servers.tsx`	View, switch, and manage configured AI servers
Add Server	`app/add-server.tsx`	Configure a new server with provider-specific fields
Edit Server	`app/edit-server.tsx`	Modify or delete existing server configuration
Backup Servers	`app/backup-servers.tsx`	Export server configurations for backup
Restore Servers	`app/restore-servers.tsx`	Import server configurations from backup
Sessions	`app/sessions.tsx`	Create, switch, and reset chat sessions (OpenClaw)
Edit Session	`app/edit-session.tsx`	Rename sessions with custom aliases
Missions	`app/missions.tsx`	View and manage multi-step AI missions (Beta)
Create Mission	`app/create-mission.tsx`	Create a new mission with goals and parameters
Mission Detail	`app/mission-detail.tsx`	View mission progress, subtasks, and streaming results
Skills	`app/skills.tsx`	Browse and install agent skill packs (OpenClaw)
Overview	`app/overview.tsx`	Gateway monitoring dashboard with health status (OpenClaw)
Gateway Logs	`app/gateway-logs.tsx`	View gateway log output (OpenClaw)
Scheduler	`app/scheduler.tsx`	Cron job management (OpenClaw)
Favorites	`app/favorites.tsx`	Saved messages viewer
Triggers	`app/triggers.tsx`	Create and manage auto-send deep links
Colors	`app/colors.tsx`	Color theme selection
Language	`app/language.tsx`	Language selection (11 languages)
Ollama Models	`app/ollama-models.tsx`	Model selection for Ollama provider
Gallery	`app/gallery.tsx`	Component showcase for development

Navigation Flow

RootLayout (ThemeProvider + KeyboardProvider + i18n)
├── OnboardingScreen (if first launch)
├── PasswordLockScreen (web, if enabled)
├── BiometricLockScreen (native, if Face ID enabled)
└── Stack Navigator
    ├── Chat (index)
    ├── Settings
    ├── Servers / Add Server / Edit Server
    ├── Backup Servers / Restore Servers
    ├── Sessions / Edit Session
    ├── Missions / Create Mission / Mission Detail
    ├── Skills
    ├── Overview / Gateway Logs (OpenClaw)
    ├── Scheduler (OpenClaw)
    ├── Favorites
    ├── Triggers
    ├── Colors
    ├── Language
    ├── Ollama Models (Ollama)
    └── Gallery

Screenshots

Getting Started

Prerequisites

Node.js 18+
pnpm package manager
Expo CLI
iOS Simulator (macOS) or Android Emulator

Quick Reference

Area	Command
Install deps	`pnpm install`
Start dev server	`pnpm start`
Run tests	`pnpm test`
Lint (auto-fix)	`pnpm lint:fix`
Format	`pnpm format`
Type check	`npx tsc --noEmit`
Build Chrome ext	`pnpm build:chrome`
Build desktop	`pnpm build:desktop`

Installation

pnpm install
pnpm start

Adding a Server

Open the app and tap the settings gear icon
Tap the + button next to "Servers"
Select a provider type
Fill in the required fields:
- OpenClaw: Name, WebSocket URL, Token, Client ID
- Claude: Name, API Key
- OpenRouter: Name, API Key
- OpenAI: Name, API Key
- OpenAI Compatible: Name, Server URL, API Key
- Gemini: Name, API Key
- Kimi: Name, API Key
- Ollama: Name, Server URL (default: http://localhost:11434)
- Apple Intelligence: Name only (on-device)
- Gemini Nano: Name only (on-device, Android)
- Echo: Name only (for testing)
Tap "Add Server"

Deep Linking

Open screens directly using lumiere:// URLs:

URL	Action
`lumiere://chat`	Chat screen
`lumiere://settings`	Settings
`lumiere://servers`	Server management
`lumiere://sessions`	Session management
`lumiere://missions`	Missions
`lumiere://overview`	Gateway overview
`lumiere://scheduler`	Cron scheduler
`lumiere://favorites`	Favorites
`lumiere://triggers`	Triggers
`lumiere://colors`	Color themes
`lumiere://language`	Language selection
`lumiere://trigger/autotrigger/{slug}`	Execute a trigger

Tech Stack

Layer	Technology
Framework	React Native 0.81, React 19, Expo 54, Expo Router 6
Language	TypeScript 5.9 (strict mode)
State management	Jotai (atoms with AsyncStorage persistence)
Navigation	Expo Router (file-based routing)
i18n	i18next + react-i18next (11 languages)
Styling	React Native StyleSheet + custom theme system
Animations	React Native Reanimated 4
Testing	Jest (jest-expo preset)
Linting	ESLint 9 (flat config) + Prettier
CI/CD	GitHub Actions, EAS Build, Fastlane, Cloudflare Pages
E2E testing	Maestro

Architecture

Project Structure

lumiere/
├── app/                    # Expo Router file-based routes (22 screens)
│   ├── _layout.tsx         # Root layout: providers, auth gates, navigation
│   ├── index.tsx           # Home screen (main chat interface)
│   ├── settings.tsx        # Settings hub
│   ├── servers.tsx         # Server management
│   ├── sessions.tsx        # Session management
│   ├── missions.tsx        # Mission list (beta)
│   ├── create-mission.tsx  # Mission creation
│   ├── mission-detail.tsx  # Mission execution view
│   ├── scheduler.tsx       # Cron job management (Molt only)
│   ├── skills.tsx          # Skills marketplace (Molt only)
│   └── ...                 # Other screens (colors, language, triggers, etc.)
│
├── src/
│   ├── components/         # React components
│   │   ├── chat/           # ChatScreen, ChatInput, ChatMessage, voice UI
│   │   ├── missions/       # Mission planning and detail UI
│   │   ├── layout/         # Sidebar, layout wrappers
│   │   ├── ui/             # Reusable UI primitives (Button, Card, Badge, Text, etc.)
│   │   ├── gallery/        # Component showcase (dev only)
│   │   └── illustrations/  # SVG illustrations
│   │
│   ├── services/           # Backend integrations and business logic
│   │   ├── providers/      # ChatProvider abstraction layer
│   │   │   ├── types.ts    # ChatProvider interface, ProviderType, capabilities
│   │   │   ├── createProvider.ts  # Factory function
│   │   │   └── CachedChatProvider.ts  # Caching wrapper
│   │   ├── molt/           # OpenClaw WebSocket client (protocol v3)
│   │   ├── claude/         # Anthropic API client
│   │   ├── openai/         # OpenAI API client
│   │   ├── openrouter/     # OpenRouter SDK wrapper
│   │   ├── gemini/         # Google Gemini API client
│   │   ├── ollama/         # Local Ollama HTTP client
│   │   ├── kimi/           # Moonshot AI Kimi client
│   │   ├── apple-intelligence/  # iOS 18+ Foundation Models
│   │   ├── gemini-nano/    # Android on-device AI
│   │   ├── echo/           # Echo test provider
│   │   ├── clawhub/        # ClawHub API integration
│   │   ├── intents/        # Chat intent system
│   │   └── notifications/  # Push/background notifications
│   │
│   ├── hooks/              # Custom React hooks (~35 hooks)
│   │   ├── useChatProvider.ts       # Provider lifecycle management
│   │   ├── useServers.ts            # Server CRUD + secure token storage
│   │   ├── useSlashCommands.ts      # 38 built-in slash commands
│   │   ├── useMissionList.ts        # Mission data (read-only)
│   │   ├── useMissionActions.ts     # Mission CRUD operations
│   │   ├── useMissionMessages.ts    # Mission message persistence
│   │   ├── useMessageQueue.ts       # Outbound message queue
│   │   ├── useAttachments.ts        # File/image attachment handling
│   │   ├── useVoiceTranscription.ts # Speech-to-text
│   │   ├── useSpeech.ts             # Text-to-speech
│   │   ├── useDeepLinking.ts        # lumiere:// URL handling
│   │   ├── useLanguage.ts           # i18n language switching
│   │   └── ...                      # Platform-specific variants (.native, .web)
│   │
│   ├── store/              # Jotai state management
│   │   ├── atoms/
│   │   │   ├── serverAtoms.ts       # Server list, current server
│   │   │   ├── sessionAtoms.ts      # Session keys, aliases
│   │   │   ├── uiAtoms.ts           # Theme, language, onboarding state
│   │   │   ├── gatewayAtoms.ts      # Connection state
│   │   │   ├── messagingAtoms.ts    # Message queue, pending share data
│   │   │   ├── userDataAtoms.ts     # Favorites, triggers
│   │   │   ├── notificationAtoms.ts # Background notification config
│   │   │   ├── missionAtoms.ts      # Mission state
│   │   │   └── secureAtom.ts        # Web-only encrypted storage
│   │   ├── storage.ts      # AsyncStorage adapter for Jotai
│   │   └── types.ts        # Store type definitions
│   │
│   ├── theme/              # Theming system
│   │   ├── ThemeContext.tsx # ThemeProvider + useTheme() hook
│   │   ├── colors.ts       # 8 color themes (lumiere, pink, green, red, blue, purple, orange, glass)
│   │   ├── themes.ts       # Light/dark mode compositions
│   │   ├── typography.ts   # Font sizes, weights, line heights
│   │   ├── spacing.ts      # Spacing scale + border radii
│   │   └── useStyles.ts    # Hook for theme-aware StyleSheets
│   │
│   ├── i18n/               # Internationalization
│   │   ├── index.ts        # i18next configuration and language detection
│   │   └── locales/        # 11 translation JSON files
│   │
│   ├── config/             # App configuration
│   │   ├── gateway.config.ts    # OpenClaw protocol v3 methods and events
│   │   └── providerOptions.tsx  # Provider-specific UI field definitions
│   │
│   ├── constants/          # Application constants
│   │   └── index.ts        # Default models, API config, cache config, session keys
│   │
│   └── utils/              # Utility functions (~23 files)
│       ├── platform.ts     # Platform detection helpers
│       ├── device.ts       # Device detection
│       ├── logger.ts       # Logging utility
│       ├── attachments.ts  # Media attachment helpers
│       └── ...             # Image/video compression, URL parsing, etc.
│
├── modules/                # Native modules (Expo autolinking)
│   ├── speech-transcription/     # iOS speech recognition
│   ├── apple-intelligence/       # iOS 18+ Foundation Models
│   ├── apple-shortcuts/          # iOS Shortcuts integration
│   └── gemini-nano/              # Android on-device Gemini
│
├── fastlane/               # iOS/Android build automation and store metadata
├── chrome-extension/       # Chrome extension assets (manifest, service worker)
├── desktop/                # Electron desktop shell
├── landing-page/           # Marketing website
├── .github/workflows/      # 13 GitHub Actions workflows
├── .maestro/               # E2E test flows (6 flows)
├── assets/                 # Images, icons, splash screens
├── public/                 # Static web assets
└── scripts/                # Build scripts (desktop, chrome extension)

Provider Abstraction

All AI backends implement the ChatProvider interface (src/services/providers/types.ts):

Interface: connect(), disconnect(), sendMessage(), getChatHistory(), resetSession(), listSessions(), getHealth()
Factory: createChatProvider(config) in src/services/providers/createProvider.ts returns the correct provider
Caching: Every provider is wrapped with CachedChatProvider for offline message persistence (max 200 messages per session via AsyncStorage)
Hook: useChatProvider() manages provider lifecycle, auto-connects on config change, and tracks connection state

Provider types: molt, claude, openai, openai-compatible, openrouter, gemini, gemini-nano, kimi, ollama, apple, echo

State Management (Jotai)

Atoms grouped by domain in src/store/atoms/
Persistence via atomWithStorage() using an AsyncStorage adapter
Web uses encrypted secureAtom (AES-GCM via Web Crypto API) for server credentials
Access outside React: getStore() from src/store/storage.ts
Session key format: agent:agentId:sessionName (mission sessions prefixed with mission-)

Theme System

ThemeProvider wraps the app, reads themeModeAtom (light/dark/system) and colorThemeAtom
8 color themes, each with light + dark variants
Use useStyles((theme) => createStyles(theme)) for theme-aware StyleSheets
Use useTheme() for direct access to theme values, mode toggling, color switching

Platform-Specific Code

Files ending in .native.ts(x) — iOS/Android only
Files ending in .web.ts(x) — web only
Default .ts(x) — fallback (usually web or shared)
Examples: KeyboardProvider, useNotifications, useQuickActions, useAppleShortcuts

Hook Composition

Large hooks are decomposed into focused hooks. Example: useMissions() composes useMissionList() + useMissionActions() + useMissionMessages().

Platform-Specific Features

iOS

Voice transcription via native Speech Recognition
Apple Intelligence provider (iOS 18+)
Apple Shortcuts integration
3D Touch quick actions
Glass effect input styling

iPad

Form sheet modals instead of full-screen
Responsive layouts
Landscape orientation support

Android

Gemini Nano on-device AI
Quick actions / app shortcuts

Web

Full browser support via Metro bundler
Password lock screen with encrypted storage

Desktop

Electron app builds (macOS, Windows, Linux)

Chrome Extension

Browser extension build with service worker

Testing

Unit Tests (Jest)

Config: jest.config.js using jest-expo preset
Test files: __tests__/ directories with .test.ts(x) or .spec.ts(x) extensions
Coverage: collects from src/**/*.{ts,tsx} (excludes .d.ts and index.ts)
Mocks: AsyncStorage and @openrouter/sdk are mocked (see __mocks__/)
Commands: pnpm test, pnpm test:watch, pnpm test:coverage

E2E Tests (Maestro)

Flows in .maestro/ directory (6 flows)
Tests onboarding, chat messaging, server addition, settings navigation, theme switching
Runs on Maestro Cloud targeting iOS (iPhone 16) and Android (API 34)

CI/CD

GitHub Actions Workflows

Workflow	Trigger	Purpose
`lint.yml`	Push/PR	ESLint + Prettier checks
`test.yml`	Push/PR	Jest with coverage
`build-check.yml`	Push/PR	TypeScript type check + Expo export
`deploy-web.yml`	Push/PR	Deploy to Cloudflare Pages
`deploy-cloudflare-pages.yml`	Push/PR (landing-page/)	Deploy landing page
`ios-build.yml`	Version tag / manual	iOS production build via EAS
`android-build.yml`	Version tag / manual	Android production build via EAS
`deploy-ios-testing.yml`	Manual	iOS testing build to TestFlight
`deploy-android-testing.yml`	Manual	Android testing build to internal track
`build-desktop.yml`	Version tag / manual	Electron builds (macOS, Win, Linux)
`export-chrome-extension.yml`	Push/PR / manual	Chrome extension build
`maestro.yml`	Manual	E2E tests on Maestro Cloud
`upload-metadata.yml`	Manual	App store metadata via Fastlane

Build Profiles (EAS)

development: Expo dev client, internal distribution
preview: Internal distribution for beta testing
production: Full app store builds with auto-submission

Internationalization

The app supports 11 languages:

Language	Code	Locale File
English	`en`	`src/i18n/locales/en.json`
Spanish	`es`	`src/i18n/locales/es.json`
French	`fr`	`src/i18n/locales/fr.json`
German	`de`	`src/i18n/locales/de.json`
Hindi	`hi`	`src/i18n/locales/hi.json`
Japanese	`ja`	`src/i18n/locales/ja.json`
Korean	`ko`	`src/i18n/locales/ko.json`
Portuguese (Brazilian)	`pt-BR`	`src/i18n/locales/pt-BR.json`
Russian	`ru`	`src/i18n/locales/ru.json`
Simplified Chinese	`zh-CN`	`src/i18n/locales/zh-CN.json`
Traditional Chinese	`zh-TW`	`src/i18n/locales/zh-TW.json`

All user-facing strings use react-i18next. Every key in en.json must exist in all other locale files. Fastlane metadata under fastlane/metadata/ must also be kept in sync.

Code Style

ESLint 9 (flat config) with @typescript-eslint and simple-import-sort plugins
Prettier: No semicolons, single quotes, trailing commas, 100-char width, 2-space indent, LF endings
TypeScript: Strict mode, extends expo/tsconfig.base

Contributing

Fork the repository
Create a feature branch (git checkout -b feature/my-feature)
Commit your changes (git commit -m 'Add my feature')
Push to the branch (git push origin feature/my-feature)
Open a Pull Request

Please run pnpm lint:fix and pnpm format before submitting.

Adding a New AI Provider

Create a new directory under src/services/<provider-name>/
Implement the ChatProvider interface from src/services/providers/types.ts
Register the provider type in the ProviderType union
Add a case in createChatProvider() in src/services/providers/createProvider.ts
Add default model constants to src/constants/index.ts
Add provider UI options to src/config/providerOptions.tsx
Add translated provider name and config labels to all 11 locale files

Adding a New Screen

Create the route file in app/<screen-name>.tsx
The root _layout.tsx auto-registers it as a modal route
Use useStyles() for theme-aware styling
All user-facing strings must use t('key') with translations in all 11 locales

Adding a New Jotai Atom

Add the atom in the appropriate file under src/store/atoms/
Use atomWithStorage() if persistence is needed
Export from src/store/index.ts
Add TypeScript types to src/store/types.ts if needed

License

MIT

Name		Name	Last commit message	Last commit date
Latest commit History 704 Commits
.github/workflows		.github/workflows
.maestro		.maestro
__mocks__		__mocks__
app		app
assets		assets
chrome-extension		chrome-extension
desktop		desktop
doc		doc
docs		docs
fastlane		fastlane
landing-page		landing-page
modules		modules
public/fonts		public/fonts
screenshots		screenshots
scripts		scripts
src		src
targets/widget		targets/widget
.gitignore		.gitignore
.prettierignore		.prettierignore
.prettierrc		.prettierrc
CLAUDE.md		CLAUDE.md
Gemfile		Gemfile
LICENSE		LICENSE
README.md		README.md
app.json		app.json
babel.config.js		babel.config.js
eas.json		eas.json
electron-builder.config.js		electron-builder.config.js
eslint.config.mjs		eslint.config.mjs
jest.config.js		jest.config.js
metro.config.js		metro.config.js
package.json		package.json
pnpm-lock.yaml		pnpm-lock.yaml
tsconfig.json		tsconfig.json

Folders and files

Latest commit

History

Repository files navigation

Lumiere

Features

Core

Voice & Input

Personalization

Automation

Security

Platform

AI Providers

OpenClaw (Molt Gateway)

Claude (Anthropic API)

OpenRouter

OpenAI

OpenAI Compatible

Gemini

Gemini Nano

Kimi

Ollama

Apple Intelligence

Echo Server

Screens

Navigation Flow

Screenshots

Getting Started

Prerequisites

Quick Reference

Installation

Adding a Server

Deep Linking

Tech Stack

Architecture

Project Structure

Provider Abstraction

State Management (Jotai)

Theme System

Platform-Specific Code

Hook Composition

Platform-Specific Features

iOS

iPad

Android

Web

Desktop

Chrome Extension

Testing

Unit Tests (Jest)

E2E Tests (Maestro)

CI/CD

GitHub Actions Workflows

Build Profiles (EAS)

Internationalization

Code Style

Contributing

Adding a New AI Provider

Adding a New Screen

Adding a New Jotai Atom

License

About

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors 4

Languages

Packages