Skip to content

habasefa/react-native-ajora

Repository files navigation

🚧 This library is under construction! πŸ› οΈ

πŸ“… It will be ready by October 15, 2025 πŸš€

πŸ€– React Native Ajora

The most complete AI agent UI for React Native

npm version License: MIT TypeScript React Native

Build beautiful, intelligent chat interfaces with AI agents in React Native


✨ Features

  • 🎨 Beautiful UI Components - Pre-built, customizable chat components with modern design
  • 🧠 AI Agent Ready - Built specifically for AI agent interactions with streaming responses
  • πŸ“± Cross-Platform - Works on iOS, Android, and Web with React Native
  • 🎯 TypeScript Support - Full TypeScript definitions for all components and props
  • πŸ”§ Highly Customizable - Extensive prop system and render functions for complete customization
  • πŸ’¬ Rich Media Support - Native support for images, audio, files with lightbox viewing
  • 🧡 Thread Management - Multi-conversation support with persistent thread history
  • ⌨️ Smart Keyboard - Intelligent keyboard handling with React Native Keyboard Controller
  • 🎭 Smooth Animations - Native animations powered by React Native Reanimated 3
  • πŸ”Œ Built-in AI Tools - Native tools including todo lists, web search, document search, and confirmations
  • πŸ“Š Thinking Indicators - Beautiful animated thinking indicators for AI processing states
  • πŸ”„ Function Calling - Full support for AI function calls with custom tool UI integration
  • πŸŽ›οΈ Server/Client Tools - Support for both server-executed and client-side interactive tools
  • πŸ’­ Thought Rendering - Display AI internal thoughts and reasoning process
  • πŸ”„ Message Streaming - Real-time streaming message updates via Server-Sent Events
  • 🎨 Action Sheets - Built-in action sheet support for message interactions

πŸš€ Quick Start

Installation

npm install react-native-ajora
# or
yarn add react-native-ajora

Peer Dependencies

Make sure you have these peer dependencies installed:

npm install @expo/vector-icons @expo/react-native-action-sheet @gorhom/bottom-sheet react-native-keyboard-controller react-native-reanimated react-native-lightbox-v2 dayjs uuid react-native-safe-area-context

Basic Usage

import React from "react";
import { SafeAreaView } from "react-native-safe-area-context";
import { Ajora, AjoraProvider } from "react-native-ajora";

const App = () => {
  return (
    <AjoraProvider
      baseUrl="http://localhost:3000" // Your backend server URL
      bearerToken="your-auth-token" // Optional authentication token
      debug={true} // Optional debug mode
    >
      <SafeAreaView style={{ flex: 1 }}>
        <Ajora
          showHeader={true}
          showThreads={true}
          isScrollToBottomEnabled={true}
          keyboardShouldPersistTaps="never"
          infiniteScroll={true}
          onSend={(messages) => {
            console.log("New messages:", messages);
          }}
        />
      </SafeAreaView>
    </AjoraProvider>
  );
};

export default App;

πŸ“– Documentation

Core Components

Ajora

The main chat component that orchestrates all other components.

<Ajora
  messages={messages}
  onSend={handleSend}
  isThinking={isThinking}
  showHeader={true}
  showThreads={true}
  renderTools={() => [TimeTool, WeatherTool]}
/>

Message Types

interface IMessage {
  _id: string;
  thread_id: string;
  role: "user" | "model";
  parts: Part[];
  created_at?: string;
  updated_at?: string;
}

interface Part {
  thought?: boolean; // Mark as AI internal thought
  functionCall?: FunctionCall; // AI tool/function call
  functionResponse?: FunctionResponse; // Tool response
  text?: string; // Text content
  audio?: {
    uri: string;
    name: string;
    size?: number;
    mimeType?: string;
  };
  image?: string; // Image URL or base64
  file?: {
    uri: string;
    name: string;
    size?: number;
    mimeType?: string;
  };
}

interface FunctionCall {
  id?: string;
  args?: Record<string, unknown>;
  name?: string;
  response?: any; // Merged response data
}

interface FunctionResponse {
  id?: string;
  name?: string;
  response?: Record<string, unknown>;
}

Advanced Features

Thread Management

import { Thread } from "react-native-ajora";

<Ajora
  showHeader={true}
  showThreads={true}
  onThreadSelect={(thread: Thread) => {
    console.log("Selected thread:", thread);
    // Thread switching is handled automatically
  }}
  onNewThread={() => {
    console.log("Creating new thread");
    // New thread creation is handled automatically
  }}
  onHeaderMenuPress={() => {
    // Custom header menu action (opens thread drawer by default)
  }}
  onHeaderPlusPress={() => {
    // Custom plus button action (creates new thread by default)
  }}
/>;

Built-in AI Tools

The library includes several native tools that work out of the box:

// These tools are automatically available:
// - todo_list: Interactive todo list management
// - confirm_action: User confirmation dialogs
// - search_web: Web search functionality
// - search_document: Document search capabilities

// The AI can use these tools automatically in function calls

Custom Tools Integration

// Create custom tool UI components
const CustomTimeTool = ({ request, submitQuery, onResponse }) => {
  const { tool } = request;
  const { timezone } = tool.args || {};

  return (
    <View>
      <Text>Getting time for: {timezone}</Text>
      {/* Your custom tool UI */}
    </View>
  );
};

<Ajora
  tools={() => [
    <CustomTimeTool key="get_current_time" toolName="get_current_time" />,
  ]}
/>;

API Service

The ApiService class provides a comprehensive interface for communicating with AI agent servers via REST endpoints and Server-Sent Events (SSE) for real-time streaming.

Configuration

import { ApiService } from "react-native-ajora";

const apiService = new ApiService({
  baseUrl: "https://your-api-server.com",
  bearerToken: "your-auth-token", // Optional
  debug: true, // Optional, enables detailed logging
});

Base URL Handling

The baseUrl you provide is intelligently processed:

  • Automatic /api appending: If your URL doesn't contain /api anywhere, it will be automatically appended

    • "https://myserver.com" β†’ "https://myserver.com/api"
    • "https://myserver.com/v1" β†’ "https://myserver.com/v1/api"
  • Preserves existing API paths: If your URL already contains /api, it remains unchanged

    • "https://myserver.com/api" β†’ "https://myserver.com/api"
    • "https://myserver.com/api/v3/agent" β†’ "https://myserver.com/api/v3/agent"
  • Trailing slash normalization: Trailing slashes are automatically removed for consistency

Streaming Responses

const cleanup = apiService.streamResponse(
  {
    type: "text",
    message: {
      _id: "1",
      role: "user",
      parts: [{ text: "Hello, AI!" }],
      createdAt: new Date(),
    },
  },
  {
    onChunk: (event) => {
      console.log("New message chunk:", event.message);
    },
    onFunctionResponse: (event) => {
      console.log("Tool response:", event.message);
    },
    onComplete: (event) => {
      console.log("Stream completed");
    },
    onError: (error) => {
      console.error("Stream error:", error);
    },
  }
);

// Clean up when done
cleanup();

Thread Management

// Get all threads
const threads = await apiService.getThreads();

// Create a new thread
const newThread = await apiService.createThread("My New Conversation");

// Get messages from a thread
const { messages, pagination } = await apiService.getMessages(
  "thread-id",
  20,
  0
);

Current Feature Support

βœ… Currently Supported:

  • Text Messages: Full support for text-based conversations
  • Tool/Function Calls: Complete integration with AI tools and function calling
  • Server-Sent Events: Real-time streaming responses
  • Thread Management: Multi-conversation support
  • Authentication: Bearer token support

🚧 Planned Support (Coming Soon):

  • Image Messages: Upload and display images in conversations
  • Audio Messages: Voice message recording and playback
  • File Attachments: Support for document and file sharing

The UI components are already built to handle these media types, but the API integration is still being developed. You can see placeholder methods in the codebase that will be activated once server-side support is ready.

Custom Styling

<Ajora
  messagesContainerStyle={{
    backgroundColor: "#f0f0f0",
  }}
  renderBubble={(props) => <CustomBubble {...props} />}
  renderInputToolbar={(props) => <CustomInputToolbar {...props} />}
  renderMessage={(props) => <CustomMessage {...props} />}
  renderMessageText={(props) => <CustomMessageText {...props} />}
  renderHeader={(props) => <CustomHeader {...props} />}
  renderThread={(props) => <CustomThread {...props} />}
  renderMessageActions={(props) => <CustomMessageActions {...props} />}
/>

Using the Hook Directly

For advanced use cases, you can use the useAjora hook directly:

import { useAjora } from "react-native-ajora";

const MyCustomChat = () => {
  const ajora = useAjora({
    baseUrl: "http://localhost:3000",
    bearerToken: "your-token",
    debug: true,
  });

  const {
    messagesByThread,
    threads,
    activeThreadId,
    isThinking,
    submitQuery,
    addNewThread,
    switchThread,
    stopStreaming,
    regenerateMessage,
  } = ajora;

  return (
    // Your custom chat UI
  );
};

Server Integration

The library expects a server that handles AI agent interactions:

// Server endpoints expected:
// GET /api/threads - Get user's conversation threads
// POST /api/threads - Create new thread
// GET /api/threads/:id/messages - Get messages for thread
// POST /api/chat - Send message and stream AI response (SSE)

// Example message format for API:
const userMessage = {
  type: "text", // or "regenerate"
  message: {
    _id: "unique-id",
    thread_id: "thread-id",
    role: "user",
    parts: [{ text: "Hello!" }],
    created_at: new Date().toISOString(),
  },
  mode: "assistant", // or "agent"
};

🎨 Screenshots

Chat Interface Beautiful chat interface with AI agent

Thread Management Multi-thread conversation management

Advanced Features Advanced AI agent features and tools

Custom Styling Customizable interface and theming

πŸ”§ Configuration

<<<<<<< Current (Your changes)

Provider Setup

import { AjoraProvider } from "react-native-ajora";

<AjoraProvider
  baseUrl="https://your-ai-server.com"
  bearerToken="your-auth-token"
  debug={false}
  initialMessages={{}} // Pre-populate messages
  initialThreads={[]} // Pre-populate threads
>
  <YourApp />
</AjoraProvider>;

Core Chat Props

=======

API Endpoint Configuration

React Native Ajora uses a flexible API service system that allows you to configure custom endpoints for your AI agent backend.

Basic API Configuration

import { ApiService } from "react-native-ajora";

// Configure your custom API service
const apiService = new ApiService({
  baseUrl: "https://your-api-domain.com",
  bearerToken: "your-auth-token", // optional
  debug: true, // optional, enables logging
});

<Ajora
  apiService={apiService}
  onSend={(messages) => {
    console.log("New messages:", messages);
  }}
/>;

Environment-Based Configuration

import { ApiService } from "react-native-ajora";

const getApiConfig = () => {
  const isDev = __DEV__;

  return {
    baseUrl: isDev
      ? "http://localhost:3000"
      : "https://your-production-api.com",
    bearerToken: process.env.API_TOKEN,
    debug: isDev,
  };
};

const apiService = new ApiService(getApiConfig());

<Ajora apiService={apiService} />;

Dynamic Configuration Updates

import { defaultApiService } from "react-native-ajora";

// Update configuration at runtime
defaultApiService.updateConfig({
  baseUrl: "https://new-endpoint.com",
  bearerToken: "new-token",
});

// Get current configuration
const currentConfig = defaultApiService.getConfig();
console.log("Current API config:", currentConfig);

Custom Headers and Authentication

const apiService = new ApiService({
  baseUrl: "https://your-api.com",
  bearerToken: "your-jwt-token",
  debug: false,
});

// The API service automatically handles:
// - Bearer token authentication
// - JSON content types
// - Error handling
// - SSE connection management

<Ajora apiService={apiService} onSend={handleSend} />;

API Endpoints Overview

Your backend should implement these endpoints:

Endpoint Method Description
/api/stream GET Server-Sent Events stream for real-time AI responses
/api/threads GET Retrieve all conversation threads
/api/threads POST Create a new conversation thread
/api/threads/:id/messages GET Get messages for a specific thread

SSE Stream Events

The streaming endpoint supports these event types:

// Events sent to your API
type UserEvent =
  | { type: "text"; message: IMessage; mode?: string }
  | { type: "function_response"; message: IMessage; mode?: string }
  | { type: "regenerate"; message: IMessage; mode?: string };

// Events received from your API
type AgentEvent =
  | { type: "message"; message: IMessage }
  | { type: "function_response"; message: IMessage }
  | { type: "thread_title"; threadTitle: string | Thread }
  | { type: "sources"; sources: SourceProps[] }
  | { type: "suggestions"; suggestions: SuggestionProps[] }
  | { type: "is_thinking"; is_thinking: boolean }
  | { type: "complete"; is_complete: boolean }
  | {
      type: "error";
      error: { thread_id: string; message_id: string; error: string };
    };

Error Handling

const apiService = new ApiService({
  baseUrl: "https://your-api.com",
  bearerToken: "your-token",
});

<Ajora
  apiService={apiService}
  onError={(error) => {
    console.error("API Error:", error);
    // Handle connection errors, auth failures, etc.
  }}
  onSend={handleSend}
/>;

Keyboard Handling

Incoming (Background Agent changes)

<Ajora
  // Thread & Header Management
  showHeader={true}
  showThreads={true}
  headerProps={{ title: "AI Assistant" }}
  threadProps={{ containerStyle: {} }}
  // Message Display
  isScrollToBottomEnabled={true}
  infiniteScroll={true}
  loadEarlier={false}
  alignTop={false}
  // Input Configuration
  placeholder="Ask me anything..."
  disableComposer={false}
  maxInputLength={2000}
  minComposerHeight={44}
  maxComposerHeight={120}
  alwaysShowSend={false}
  // Media & Interaction
  showUserAvatar={false}
  showAvatarForEveryMessage={false}
  renderAvatarOnTop={false}
  isCustomViewBottom={false}
/>

Keyboard Handling

<Ajora
  keyboardShouldPersistTaps="never" // "always" | "never" | "handled"
  focusOnInputWhenOpeningKeyboard={true}
  isKeyboardInternallyHandled={true}
  bottomOffset={0} // Distance from screen bottom
/>

Event Callbacks

<Ajora
  onSend={(messages) => console.log("Sent:", messages)}
  onLoadEarlier={() => console.log("Loading earlier messages")}
  onPress={(context, message) => console.log("Message pressed")}
  onLongPress={(context, message) => console.log("Message long pressed")}
  onInputTextChanged={(text) => console.log("Input changed:", text)}
  onPressActionButton={() => console.log("Action button pressed")}
/>

πŸ› οΈ Development

Prerequisites

  • Node.js >= 18
  • React Native development environment
  • Expo CLI (for example app)

Setup

  1. Clone the repository:
git clone https://github.com/habasefa/react-native-ajora.git
cd react-native-ajora
  1. Install dependencies:
npm install
  1. Start the example app:
npm start

Building

npm run build

Testing

npm test
npm run test:coverage

πŸ“š Examples

Check out the included example applications:

Example App (example-app/)

  • Complete Expo App - Full React Native app with Ajora integration
  • Native Tools Demo - Showcases built-in AI tools (todo, search, confirm)
  • Thread Management - Multi-conversation interface
  • Custom Styling - Themed components and layouts
  • Media Support - Image, audio, and file message examples

Example Server (example-server/)

  • Node.js Backend - Complete server implementation for AI agents
  • Gemini AI Integration - Example using Google's Gemini AI
  • Tool Execution - Server-side tool handling
  • SSE Streaming - Real-time message streaming
  • Database Integration - Message and thread persistence

To run the examples:

# Install dependencies
npm install

# Start the example server
cd example-server
npm install && npm start

# In another terminal, start the example app
cd example-app
npm install && npm start

🀝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

  1. Fork the repository
  2. Create your feature branch (git checkout -b feature/amazing-feature)
  3. Commit your changes (git commit -m 'Add some amazing feature')
  4. Push to the branch (git push origin feature/amazing-feature)
  5. Open a Pull Request

πŸ“„ License

This project is licensed under the MIT License - see the LICENSE file for details.

πŸ“‹ API Reference

Key Exports

// Main Components
import {
  Ajora, // Main chat component
  AjoraProvider, // Context provider
  useAjora, // State management hook
} from "react-native-ajora";

// UI Components
import {
  Actions, // Action button component
  Bubble, // Message bubble component
  Composer, // Text input component
  InputToolbar, // Input toolbar container
  LoadEarlier, // Load earlier button
  Message, // Individual message component
  MessageContainer, // Messages list container
  MessageImage, // Image message component
  MessageText, // Text message component
  Send, // Send button component
} from "react-native-ajora";

// Types
import type {
  IMessage, // Core message interface
  Part, // Message part interface
  FunctionCall, // Function call interface
  FunctionResponse, // Function response interface
  Thread, // Thread interface
  AjoraProps, // Main component props
  AjoraState, // Hook state type
} from "react-native-ajora";

Native Tools

// Built-in tools available to AI agents:
-"todo_list" - // Interactive todo list management
  "confirm_action" - // User confirmation dialogs
  "search_web" - // Web search functionality
  "search_document"; // Document search capabilities

πŸ™ Acknowledgments

  • Built with ❀️ for the React Native community
  • Inspired by modern chat interfaces and AI agent UIs
  • Special thanks to Farid and react-native-giftedchat contributors
  • Powered by React Native Reanimated, Expo Vector Icons, and other amazing open source libraries

πŸ“ž Support


About

The most complete AI agent UI for React Native

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published