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

fix_character_prompts.py

@@ -0,0 +1,65 @@
+#!/usr/bin/env python3
+"""
+Fix character system prompts to use proper template format
+"""
+
+import asyncio
+from datetime import datetime, timezone
+from sqlalchemy import select
+from src.database.connection import init_database, get_db_session
+from src.database.models import Character
+
+PROPER_SYSTEM_PROMPT_TEMPLATE = """You are a character named {{name}}. You have the following personality: {{personality}}
+
+Your speaking style is {{speaking_style}}. You are interested in {{interests}}.
+
+Background: {{background}}
+
+When responding to messages:
+1. Stay in character at all times
+2. Reference your personality and interests naturally
+3. Engage authentically with other characters
+4. Show growth and development over time
+
+Remember to be consistent with your established personality while allowing for natural character development through interactions."""
+
+async def fix_character_prompts():
+    """Fix all character system prompts to use proper template format"""
+    
+    await init_database()
+    
+    async with get_db_session() as session:
+        # Get all characters
+        characters_query = select(Character)
+        characters = await session.scalars(characters_query)
+        
+        updated_count = 0
+        
+        for character in characters:
+            print(f"\nChecking character: {character.name}")
+            print(f"Current system prompt length: {len(character.system_prompt or '') if character.system_prompt else 0}")
+            
+            # Check if the prompt needs fixing (doesn't contain template variables)
+            current_prompt = character.system_prompt or ""
+            
+            # If it doesn't contain template variables or is just raw personality text, fix it
+            if "{{name}}" not in current_prompt or len(current_prompt) < 100:
+                print(f"  - Fixing system prompt for {character.name}")
+                
+                # Use the proper template
+                character.system_prompt = PROPER_SYSTEM_PROMPT_TEMPLATE
+                character.updated_at = datetime.now(timezone.utc)
+                
+                updated_count += 1
+                print(f"  - Updated!")
+            else:
+                print(f"  - System prompt looks good, skipping")
+        
+        if updated_count > 0:
+            await session.commit()
+            print(f"\n✅ Successfully updated {updated_count} character(s)")
+        else:
+            print(f"\n✅ All characters already have proper system prompts")
+
+if __name__ == "__main__":
+    asyncio.run(fix_character_prompts())