Implement storage layer and adaptive learning tools for TutorX-MCP

Refactor response handling in app.py and various tool modules to improve JSON parsing and error management. Introduce utility functions for cleaning and extracting JSON from text, ensuring consistent data handling across quiz, lesson, learning path, and interaction tools. Enhance robustness by accommodating multiple response formats, including strings and dictionaries, to streamline educational interactions.

Refactor TutorX MCP server to integrate Mistral OCR for document processing, update concept graph tools for LLM-driven responses, and enhance learning path generation with Gemini. Transitioned various tools to utilize LLM for improved educational interactions and streamlined API responses.

Refactor TutorX MCP server to remove legacy client and utility modules, update app.py for SSE integration, and enhance .gitignore to exclude .cursor directory. Clean up main.py for improved server configuration and streamline run script for better usability.