Feat/stream event for ai sdk - #15

Description

🔄 Streaming Methods

• Added stream() method for step-by-step agent execution visibility
• Added streamEvents() method for token-level streaming with fine-grained
  event tracking

🚀 AI SDK Integration

• Added utilities for seamless integration with Vercel AI SDK (ai package)
• New utility functions: streamEventsToAISDK(),
  streamEventsToAISDKWithTools(), and createReadableStreamFromGenerator()
• Full compatibility with React hooks like useCompletion and useChat

📚 Enhanced Documentation

• Comprehensive API documentation for all streaming methods
• Detailed examples showing differences between run(), stream(), and
  streamEvents()
• Complete AI SDK integration guide with Next.js examples

🧪 Examples & Tests

• Added stream_example.ts demonstrating both streaming methods
• Added stream_events_example.ts with comprehensive streamEvents usage

Technical Details:

The implementation provides three complementary methods:

• run(): Simple execution with final result only
• stream(): Step-by-step execution with intermediate tool usage
• streamEvents(): Token-level streaming with full LangChain event tracking

The AI SDK integration enables developers to build streaming UIs that can
display:

• Real-time token-by-token responses
• Tool usage notifications
• Progress indicators
• Error handling

This makes the library suitable for production chat applications, debugging
tools, and any scenario requiring real-time agent execution visibility.