Refactor to centralize Reviewer_Engine and add LLM-based skill extraction

[MalakHisham121]

This pull request introduces significant improvements to the Reviewer Recommender system, focusing on enhanced architecture, smart caching, richer API contracts, and improved modularity. The documentation (ReviewerRecommender.md) has been extensively rewritten to clarify the new multi-agent, cache-aware workflow and the structure of the core modules. In the code, legacy methods have been removed, and normalization of required reviewers has been improved for better Jira integration.

Key changes include:

Architecture and Documentation Overhaul

• The system architecture is now described as a "Smart Caching Architecture," detailing the multi-agent workflow, cache layers, and integration with local LLMs and Jira. The process flow and API contracts are rewritten for clarity, and the active module structure is documented with clear roles for each Python file.

API and Data Handling Improvements

• The API contract for reviewer recommendations now supports richer input for required reviewers, including custom jira_username and raw_skills fields, and provides detailed request/response shapes for both /api/recommend/v2 and /api/orchestrator endpoints.
• The normalization logic for required reviewers in Reviewer_Engine.py is updated to consistently handle both string and dictionary forms, ensuring jira_username is captured and available for downstream processing.

Codebase Simplification and Modernization

• Legacy, unused methods for processing specific and top developers, as well as deprecated initialization methods, are removed from Reviewer_Engine.py, reducing technical debt and focusing on the smart caching workflow. [1] [2]
• The _initialize_specific_developers method is improved to handle both string and dict forms for required developers, extracting usernames robustly.

Modularization and Agent Integration

• The orchestrator now explicitly imports the analyze_jira_context agent, reflecting the modular, multi-agent architecture described in the documentation.

These changes collectively modernize the reviewer recommendation system, making it more efficient, maintainable, and ready for integration with advanced AI and organizational workflows.…iewer_Engine implementation and add LLM-based skill extraction for commit history.