AI/LLM Integration & RAG System

Major features

• LLM and RAG integration. Full semantic intelligence suite powered by large language models.
• LLM orchestration. Llama 3 via Groq for fast, cost-efficient inference.
• ChromaDB. Semantic code indexing for context-aware analysis across your repository.
• refactron suggest. AI-powered refactoring suggestions with explanations and risk scores.
• refactron document. Automated docstring generation for functions, classes, and modules.
• Repository management. New commands: refactron repo list, repo connect, and repo disconnect.
• Observability. refactron metrics for performance stats, serve-metrics for a Prometheus endpoint, and refactron telemetry for opt-in analytics.
• CI/CD integration. refactron ci generates GitHub Actions and GitLab CI configuration files.

Bug fixes

• URL sanitization. Security-hardened URL sanitization to prevent injection attacks.
• Python 3.8. Resolved compatibility issues on Python 3.8 environments.
• GroqClient. Fixed missing GroqClient dependency causing import errors on fresh installs.