Download the PHP package aitoolkit/laravel-ai-toolkit without Composer

Laravel AI Toolkit

A professional, production-ready Laravel package for integrating multiple AI providers with advanced features like async processing, real-time streaming, intelligent caching, and error handling.

✨ Features

🤖 Multi-Provider AI Support

• OpenAI - GPT-4, GPT-3.5, and embeddings
• Anthropic - Claude 4 Sonnet, Claude 3 Opus, and more
• Groq - Ultra-fast inference with Mixtral and Llama models
• Easy provider switching via configuration

⚡ Performance & Reliability

• Async Queue Jobs - Non-blocking AI operations
• Intelligent Caching - Redis/Database cache with TTL and invalidation
• Circuit Breaker Pattern - Automatic failure protection
• Exponential Backoff - Smart retry logic with jitter
• Rate Limiting - Built-in API rate limit handling

🔄 Real-Time Features

• Streaming Responses - Real-time AI response streaming
• Laravel Reverb Integration - WebSocket broadcasting for live updates
• Event-Driven Architecture - Dispatch events for UI updates

🔧 Developer Experience

• Type-Safe Contracts - Interface-driven provider architecture
• Comprehensive CLI - Command-line tools for testing and management
• Laravel Facades - Easy dependency injection
• Rich Configuration - Environment-based provider settings
• Extensible Design - Easy to add new providers

📦 Installation

Requirements

• PHP 8.3 or higher
• Laravel 11 or 12
• Composer

Install via Composer

Publish Configuration

This creates config/ai-toolkit.php where you can configure your providers.

Environment Variables

Add your API keys to .env:

🚀 Quick Start

Basic Usage

Using Facades

Async Jobs

Streaming Responses

CLI Usage

⚙️ Configuration

Provider Settings

Cache Settings

Queue Settings

Broadcasting Settings

🏗️ Architecture

Provider Pattern

All AI providers implement AIProviderContract:

Service Architecture

Caching Strategy

Retry Logic

🧪 Testing

Run the test suite:

Run with coverage:

The package includes:

• Unit Tests - Individual component testing
• Feature Tests - Integration testing
• Provider Tests - Mock API testing
• Service Tests - Caching, retry logic, and jobs

📚 API Reference

AIProviderContract

chat(string $prompt, array $options = []): array

Send a chat message to the AI provider.

Parameters:

• $prompt - The user message
• $options - Additional parameters (model, max_tokens, temperature, etc.)

Returns:

stream(string $prompt, array $options = []): StreamedResponse

Stream a chat response in real-time.

embed(string $text): array

Generate text embeddings (OpenAI only).

AiCacheService

remember(string $operation, string $provider, string $input, callable $callback, array $options = []): mixed

Cache AI responses with automatic TTL.

generateKey(string $operation, string $provider, string $input, array $options = []): string

Generate consistent cache keys.

invalidatePattern(string $pattern): int

Invalidate cache by pattern.

RetryService

execute(callable $callback, array $options = []): mixed

Execute operations with retry logic.

Options:

• max_retries - Maximum retry attempts (default: 3)
• base_delay - Base delay in seconds (default: 1)
• strategy - 'exponential', 'linear', or 'fixed'
• circuit_breaker - Enable circuit breaker (default: true)

🔒 Security

• API Key Validation - Validates provider keys on initialization
• Input Sanitization - All inputs are sanitized before API calls
• Rate Limiting - Built-in rate limit handling
• Error Handling - No sensitive data in error messages
• Caching - Cached responses don't include API keys

🚀 Performance Tips

1. Enable Caching

2. Use Async Jobs for Long Operations

3. Monitor Cache Hit Rates

4. Use Circuit Breakers

📖 Advanced Usage

Custom Providers

Create a custom provider by implementing the contract:

Register in your service provider:

Event Listeners

Custom Caching

🐛 Troubleshooting

Common Issues

1. "Invalid AI provider" Error

• Check your ai.default_provider configuration
• Ensure the provider name matches exactly: 'openai', 'anthropic', or 'groq'

2. API Key Not Found

• Verify environment variables are set
• Check for typos in variable names
• Ensure the config file is published

3. Streaming Not Working

• Enable broadcasting in config
• Set up Laravel Reverb or Pusher
• Check browser console for WebSocket errors

4. Queue Jobs Not Running

• Ensure queue driver is configured
• Run php artisan queue:work
• Check Laravel logs for errors

Debug Mode

Enable detailed logging:

🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Write tests for new functionality
4. Ensure all tests pass
5. Submit a pull request

Development Setup

📄 License

This package is open-sourced software licensed under the MIT license.

🆘 Support

• 🐛 Issues: GitHub Issues

🙏 Acknowledgments

• OpenAI for their excellent APIs
• Anthropic for Claude
• Groq for lightning-fast inference
• Laravel for the amazing framework

🎛️ Filament Admin Panel Integration

The Laravel AI Toolkit includes optional Filament Admin Panel integration for a beautiful web interface to manage AI providers, monitor usage, and test AI operations.

Quick Setup

Features

• ✅ AI Provider Management - Configure multiple providers (OpenAI, Anthropic, Groq)
• ✅ Real-time Chat Dashboard - Interactive AI testing interface
• ✅ Usage Analytics - Monitor requests, cache hit rates, response times
• ✅ Provider Testing - Built-in connection testing for all providers
• ✅ Encrypted Storage - API keys encrypted in database
• ✅ Dynamic Configuration - Database-driven provider management

Database-Driven Architecture

The Filament integration uses a database-first approach:

1. Providers stored in database (not config files)
2. Encrypted API keys for security
3. Dynamic configuration via admin UI
4. Seeder creates initial providers from .env values

Documentation

📖 Full documentation available in: docs/FILAMENT_INTEGRATION.md

Includes detailed setup, configuration, troubleshooting, and advanced usage examples.

Made with ❤️ for the Laravel community