refactor: enhance LLM configuration and initialization

[konvar]

BREAKING CHANGE: Complete overhaul of LLM configuration and initialization system

Core Changes:

• Add dynamic model selection for all LLM backends
• Implement availability checking for all LLM types

Features:

• Add color-coded menu for LLM backend selection
• Show setup instructions for unconfigured backends
• Add real-time status checks for all LLM types
• Implement proper error handling and user feedback

Configuration:

• Clear Gemini API key
• Reset Ollama model name
• Add proper validation for model paths and API keys

Security:

• Add checks for default/placeholder values
• Improve error messages for missing credentials

Dependencies:

• Add google-generativeai>=0.3.0 to requirements.txt

[konvar]

I thought I was modifying https://github.com/TheBlewish/Automated-AI-Web-Researcher-Ollama 🤣 only after I finished I did realize that I had modified an older project.
I wanted to add Gemini after seeing a Reddit comment saying that Gemini has a large context window and I realized he was right. I hope you can copy or get inspired 😢