feat: implement ReasoningAgent and DualBrainAgent with advanced reasoning capabilities

[github-actions]

Implements ReasoningAgent and DualBrainAgent classes as requested in issue #968

Changes:

• Add ReasoningConfig class for configurable reasoning parameters
• Add ActionState enum for flow control
• Implement ReasoningAgent inheriting from Agent with step-by-step reasoning
• Implement DualBrainAgent with dual-LLM architecture for analysis vs conversation
• Add confidence scoring integration and reasoning trace tracking
• Maintain full backward compatibility with existing Agent API

Features:

• Step-by-step reasoning with confidence scoring
• Separate LLMs for conversation and reasoning
• Reasoning trace tracking with start/complete lifecycle
• Action states for flow control
• Thread-safe LLM switching in DualBrainAgent
• All API examples from issue Create ReasoningAgent inherited from Agent clas... #968 work as specified

Architecture Compliance:

• Follows AGENTS.md guidelines with protocol-driven design
• Uses lazy imports for performance
• Centralized logging patterns
• Backward compatible with existing Agent class
• Proper error handling and validation

Fixes from Original PR #977:

• Fixed thread-safety issues with LLM switching
• Proper error handling when LLM returns None
• Removed unused imports and parameters
• Fixed step counting logic for reasoning traces
• Added comprehensive test suite

Testing:

All tests pass including:

• Import functionality
• Reasoning module components
• ReasoningAgent initialization and methods
• DualBrainAgent dual-brain functionality

Resolves #968

🤖 Generated with Claude Code

[MervinPraison]

Replaced by #1227 to properly attribute authorship to MervinPraison.