genai-hooks

genai-hooks

Production-ready React hooks for OpenAI integration with full TypeScript support, secure API key management, and comprehensive error handling.

Features

• 🚀 Production-Ready: Built with enterprise-grade patterns and best practices
• 🔐 Secure API Management: Support for backend proxies to protect API keys
• 📝 Full TypeScript Support: Complete type definitions for excellent DX
• 🔄 Advanced Error Handling: Automatic retries, cancellation, and error recovery
• ⚡ Performance Optimized: Tree-shakeable, minimal bundle size
• 🧪 Well Tested: Comprehensive test coverage
• 🎯 Modern React: Supports React 16.8+, 17, and 18

Installation

npm install genai-hooks
# or
yarn add genai-hooks
# or
pnpm add genai-hooks

Quick Start

Basic Usage (Development Only)

import { AIHooksProvider, useTextGeneration } from 'genai-hooks';

function App() {
  return (
    <AIHooksProvider config={{ apiKey: process.env.REACT_APP_OPENAI_API_KEY }}>
      <TextGenerator />
    </AIHooksProvider>
  );
}

function TextGenerator() {
  const { generateText, data, isLoading, error } = useTextGeneration();
  
  const handleGenerate = () => {
    generateText('Write a short poem about React hooks');
  };

  return (
    <div>
      <button onClick={handleGenerate} disabled={isLoading}>
        {isLoading ? 'Generating...' : 'Generate Text'}
      </button>
      {error && <p>Error: {error.message}</p>}
      {data && <p>{data}</p>}
    </div>
  );
}

Production Usage (Recommended)

For production applications, use a backend proxy to secure your API keys:

import { AIHooksProvider } from 'genai-hooks';

function App() {
  return (
    <AIHooksProvider config={{ 
      baseURL: '/api/ai', // Your backend proxy endpoint
      headers: {
        'Authorization': `Bearer ${userToken}` // Your auth token
      }
    }}>
      <YourApp />
    </AIHooksProvider>
  );
}

Available Hooks

useTextGeneration

Generate text using OpenAI's completion models.

const {
  generateText,  // Function to trigger generation
  data,          // Generated text
  isLoading,     // Loading state
  error,         // Error object
  cancel,        // Cancel ongoing request
  reset          // Reset hook state
} = useTextGeneration();

// With options
generateText('Your prompt', {
  model: 'gpt-3.5-turbo-instruct',
  maxTokens: 150,
  temperature: 0.7,
  topP: 1,
  frequencyPenalty: 0,
  presencePenalty: 0,
  stop: ['\n']
});

useImageGeneration

Generate images using DALL-E.

const {
  generateImage,
  data,         // Image URL or base64 data
  isLoading,
  error,
  cancel,
  reset
} = useImageGeneration();

// With options
generateImage('A futuristic city at sunset', {
  model: 'dall-e-3',
  size: '1024x1024',
  quality: 'hd',
  style: 'vivid',
  n: 1,
  responseFormat: 'url' // or 'b64_json'
});

usePredictiveCompletion

Get real-time text suggestions as users type.

const {
  suggestions,      // Array of suggestions
  isLoading,
  error,
  fetchSuggestions, // Debounced function
  cancel,
  reset
} = usePredictiveCompletion({
  maxSuggestions: 3,
  debounceMs: 300,
  temperature: 0.7
});

// In your input handler
const handleInputChange = (e) => {
  setText(e.target.value);
  fetchSuggestions(e.target.value);
};

useLanguageTranslation

Translate text between languages.

const {
  translateText,
  data,          // Translated text
  isLoading,
  error,
  cancel,
  reset
} = useLanguageTranslation();

// Translate
translateText('Hello world', 'Spanish', {
  model: 'gpt-3.5-turbo',
  temperature: 0.3
});

Configuration

Provider Configuration

<AIHooksProvider config={{
  // API Configuration
  apiKey: string,           // OpenAI API key (use env vars)
  baseURL: string,          // API base URL (default: OpenAI)
  
  // Request Configuration  
  timeout: number,          // Request timeout in ms (default: 30000)
  retries: number,          // Number of retries (default: 3)
  retryDelay: number,       // Delay between retries (default: 1000)
  
  // Custom headers
  headers: Record<string, string>,
  
  // Error handler
  onError: (error: AIHookError) => void
}}>

Dynamic Configuration

function App() {
  const { updateConfig } = useAIHooksConfig();
  
  // Update configuration at runtime
  const switchToProduction = () => {
    updateConfig({
      baseURL: 'https://api.mycompany.com/ai',
      headers: { 'X-API-Version': 'v2' }
    });
  };
}

Security Best Practices

1. Backend Proxy (Recommended)

Never expose your OpenAI API key in client-side code. Use a backend proxy:

// Express.js example
app.post('/api/ai/completions', authenticate, async (req, res) => {
  const response = await openai.completions.create({
    ...req.body,
    // Override any sensitive fields
    api_key: process.env.OPENAI_API_KEY
  });
  res.json(response);
});

2. Environment Variables

If you must use client-side API keys (development only):

# .env.local
REACT_APP_OPENAI_API_KEY=your_api_key_here

3. Rate Limiting

Implement rate limiting on your backend:

const rateLimit = require('express-rate-limit');

const limiter = rateLimit({
  windowMs: 15 * 60 * 1000, // 15 minutes
  max: 100 // limit each IP to 100 requests per windowMs
});

app.use('/api/ai', limiter);

Advanced Usage

Custom Error Handling

<AIHooksProvider config={{
  onError: (error) => {
    // Log to error tracking service
    console.error('AI Hook Error:', error);
    
    // Show user-friendly message
    if (error.statusCode === 429) {
      toast.error('Rate limit exceeded. Please try again later.');
    }
  }
}}>

Request Cancellation

function InterruptibleGenerator() {
  const { generateText, cancel, isLoading } = useTextGeneration();
  
  return (
    <>
      <button onClick={() => generateText('Long prompt...')}>
        Generate
      </button>
      {isLoading && (
        <button onClick={cancel}>Cancel</button>
      )}
    </>
  );
}

Streaming Responses (Coming Soon)

const { streamText, data, isStreaming } = useTextStream();

streamText('Write a story...', {
  onChunk: (chunk) => console.log('Received:', chunk),
  onComplete: (fullText) => console.log('Complete:', fullText)
});

TypeScript

All hooks are fully typed. Import types as needed:

import type { 
  TextGenerationOptions,
  ImageGenerationOptions,
  AIHookError,
  AIHooksConfig 
} from 'genai-hooks';

Error Handling

The library provides detailed error information:

interface AIHookError {
  message: string;
  code?: string;
  statusCode?: number;
  details?: unknown;
}

Common error codes:

• 429: Rate limit exceeded
• 401: Invalid API key
• 500: Server error
• ECONNABORTED: Request timeout

Contributing

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

License

MIT © Harsh Joshi

Support

• 📚 Documentation
• 🐛 Issue Tracker
• 💬 Discussions