Implement comprehensive LLM provider system with global cost protection

- Add multi-provider LLM architecture supporting OpenRouter, OpenAI, Gemini, and custom providers
- Implement global LLM on/off switch with default DISABLED state for cost protection
- Add per-character LLM configuration with provider-specific models and settings
- Create performance-optimized caching system for LLM enabled status checks
- Add API key validation before enabling LLM providers to prevent broken configurations
- Implement audit logging for all LLM enable/disable actions for cost accountability
- Create comprehensive admin UI with prominent cost warnings and confirmation dialogs
- Add visual indicators in character list for custom AI model configurations
- Build character-specific LLM client system with global fallback mechanism
- Add database schema support for per-character LLM settings
- Implement graceful fallback responses when LLM is globally disabled
- Create provider testing and validation system for reliable connections

requirements.txt

@@ -35,4 +35,5 @@ python-jose[cryptography]>=3.3.0
 passlib[bcrypt]>=1.7.4
 websockets>=12.0
 psutil>=5.9.6
-python-socketio>=5.10.0
+python-socketio>=5.10.0,<6.0.0
+python-engineio>=4.7.0,<5.0.0