feat: Implement routing pattern with native framework implementations

[lkronecker13]

Summary

Implement comprehensive routing pattern demonstrations across three major AI frameworks (LangChain, LlamaIndex, Google ADK) with proper framework-native architectures and extensive documentation.

Key Features

• LangChain Routing: Uses RunnableBranch for conditional execution with complex data flow
• LlamaIndex Routing: Native RouterQueryEngine with CustomQueryEngine classes and PydanticSingleSelector
• Google ADK Routing: Sub-agent architecture with Auto-Flow delegation patterns
• Comprehensive Documentation: 200+ line README explaining routing theory, mechanisms, and applications
• LLM-Based Intelligence: All implementations use intelligent LLM-driven routing decisions rather than hardcoded logic

Technical Implementation

LangChain (langchain_routing.py)

• Uses RunnableBranch for conditional routing logic
• Demonstrates complex data passing between routing and execution phases
• Framework-integrated approach with proper LangChain patterns

LlamaIndex (llamaindex_routing.py)

• Native RouterQueryEngine with PydanticSingleSelector for intelligent routing
• Custom Query Engines: BookingQueryEngine, InfoQueryEngine, UnclearQueryEngine extending CustomQueryEngine
• QueryEngineTool abstractions with detailed descriptions for optimal routing decisions
• Refactored from manual routing to proper framework architecture

Google ADK (google_adk_routing.py)

• Enterprise-grade sub-agent delegation pattern
• Built-in Auto-Flow routing with coordinator-sub-agent architecture
• Function tool integration for specialized handlers

Routing Scenarios Tested

All implementations handle three test scenarios:

1. Booking Request: "Book me a flight to London" → Booking handler
2. Information Query: "What is the capital of Italy?" → Info handler
3. Complex/Unclear: "Tell me about quantum physics" → Unclear handler

Quality Assurance

• ✅ Type Safety: Full MyPy compliance across all implementations
• ✅ Code Quality: Ruff formatting and linting passed
• ✅ Test Coverage: 96% coverage maintained, all 21 tests passing
• ✅ Documentation: Comprehensive README with theory, examples, and best practices
• ✅ Make Targets: Proper integration with project build system

Architecture Improvements

• Framework-Native Patterns: Each implementation uses its framework's intended routing architecture
• Intelligent Decision Making: LLM-based routing provides contextual awareness over hardcoded logic
• Extensible Design: Easy to add new routing categories or handlers
• Production Ready: Proper error handling, type safety, and documentation

Test Plan

• All three routing implementations execute successfully
• LLM routing decisions are contextually appropriate
• Debug output shows proper handler delegation
• Type checking passes with no errors
• All unit tests continue to pass
• Make targets work correctly for all routing examples
• Documentation accurately reflects implementations

🤖 Generated with Claude Code