[Feat]Ralph Loop Implementation with CLI Agent Orchestrator (CAO)

## Overview
Implement Ralph Loop functionality within CLI Agent Orchestrator (CAO) to enable iterative, adaptive task execution with support for multiple CLI providers including Kiro-CLI, Claude Code, and Gemini CLI.

## User Stories
- As a developer, I want to use Ralph Loop for iterative problem-solving so that complex tasks can be broken down and executed incrementally
- As a CLI Agent Orchestrator user, I want Ralph Loop to work seamlessly with Kiro-CLI, Claude Code, and Gemini CLI so that I can choose the best tool for each iteration
- As a power user, I want the loop to adapt based on intermediate results so that my workflows can handle dynamic scenarios
- As a team lead, I want a standardized loop implementation so that my team can build consistent, maintainable automation workflows

## Acceptance Criteria
- Ralph Loop core architecture is implemented and functional
- Ralph Loop successfully executes tasks using Kiro-CLI
- Ralph Loop successfully executes tasks using Claude Code
- Loop iteration logic handles success, failure, and retry scenarios
- Loop can maintain context across iterations
- Documentation includes Ralph Loop architecture, usage examples, and integration patterns
- Unit tests cover core loop functionality and task execution
- Integration tests validate end-to-end workflows with multiple providers

## Proposed solution

## Additional context
**Is your feature request related to a problem? Please describe.**
Complex automation tasks often require iterative problem-solving with feedback loops and the ability to adapt based on intermediate results. Without a structured loop implementation, users must manually orchestrate iterations, leading to brittle workflows that are difficult to maintain and extend. The lack of a standardized loop pattern means users repeatedly implement custom iteration logic instead of leveraging a proven, reusable framework.

**Describe alternatives you've considered**
- Implementing custom loops in user scripts (lacks standardization and reusability)
- Using existing workflow engines (may not integrate well with CLI Agent Orchestrator's architecture)
- Building a simpler iteration mechanism without adaptive capabilities (reduces flexibility for complex scenarios)
- Hardcoding provider selection instead of dynamic routing (limits the range of capabilities available to the loop)
- Using only a single CLI provider (limits the range of capabilities and reduces resilience)

**Additional context**
- The loop should integrate with the existing CAO architecture and leverage the hybrid workflow capabilities
- Consider leverage Skills as a integration (optional)