feat: Add comprehensive retry and resilience system for production reliability

[sarvottam-bhagat]

🛡️ Production-Ready Resilience System

Overview

This PR adds a comprehensive retry and resilience system to the Langbase SDK, making it production-ready for enterprise applications.

✨ Features Added

Core Resilience Infrastructure

• Multiple Retry Strategies: Exponential, Linear, Fixed, Fibonacci backoff
• Circuit Breakers: Prevent cascade failures with configurable thresholds
• Rate Limit Handling: Automatic respect for Retry-After headers
• Jitter Support: Prevents thundering herd problems
• Smart Error Detection: Configurable retry conditions

Integration & Compatibility

• Seamless Integration: Works with existing Request system
• Backward Compatible: Existing code continues to work unchanged
• Easy Configuration: Simple API with sensible defaults
• Streaming Support: Works with streaming responses

🧪 Testing

• 20+ comprehensive tests covering all retry strategies
• Circuit breaker state transitions and recovery scenarios
• Integration tests with actual Request class
• Edge cases and error conditions

📚 Documentation

• Complete documentation in docs/resilience.md
• Working examples in examples/resilience/
• Updated README with resilience features

🎯 Use Cases

• E-commerce platforms handling traffic spikes
• Financial applications requiring high reliability
• Content generation pipelines processing large volumes
• Customer support chatbots needing 24/7 uptime

💡 Business Impact

• Reduces support burden (fewer failure-related tickets)
• Improves user experience (more reliable applications)
• Enables enterprise adoption (production-ready reliability)
• Competitive advantage (matches best-in-class SDKs)