Enhance AICore model validation and logging

- Improved model configuration handling in AICore to ensure only supported models are set as default.
- Added error logging for invalid model configurations and warnings for missing models.
- Implemented validation checks for the active model to provide clear error messages when no valid model is configured.
- Enhanced user feedback during AI model selection to improve debugging and configuration clarity.

backend/igny8_core/ai/ai_core.py

@@ -58,8 +58,17 @@ class AICore:
                 if openai_settings and openai_settings.config:
                     self._openai_api_key = openai_settings.config.get('apiKey')
                     model = openai_settings.config.get('model')
-                    if model and model in MODEL_RATES:
-                        self._default_model = model
+                    if model:
+                        if model in MODEL_RATES:
+                            self._default_model = model
+                            logger.info(f"Loaded model '{model}' from IntegrationSettings for account {self.account.id}")
+                        else:
+                            error_msg = f"Model '{model}' from IntegrationSettings is not in supported models list. Supported models: {list(MODEL_RATES.keys())}"
+                            logger.error(f"[AICore] {error_msg}")
+                            logger.error(f"[AICore] Account {self.account.id} has invalid model configuration. Please update Integration Settings.")
+                            # Don't set _default_model, will fall back to Django settings
+                    else:
+                        logger.warning(f"No model configured in IntegrationSettings for account {self.account.id}, will use fallback")
                 
                 # Load Runware settings
                 runware_settings = IntegrationSettings.objects.filter(
@@ -148,6 +157,38 @@ class AICore:
         
         # Step 2: Determine model
         active_model = model or self._default_model
+        
+        # Validate model is available and supported
+        if not active_model:
+            error_msg = 'No AI model configured. Please configure a model in Integration Settings or Django settings.'
+            logger.error(f"[AICore] {error_msg}")
+            tracker.error('ConfigurationError', error_msg)
+            return {
+                'content': None,
+                'error': error_msg,
+                'input_tokens': 0,
+                'output_tokens': 0,
+                'total_tokens': 0,
+                'model': None,
+                'cost': 0.0,
+                'api_id': None,
+            }
+        
+        if active_model not in MODEL_RATES:
+            error_msg = f"Model '{active_model}' is not supported. Supported models: {list(MODEL_RATES.keys())}"
+            logger.error(f"[AICore] {error_msg}")
+            tracker.error('ConfigurationError', error_msg)
+            return {
+                'content': None,
+                'error': error_msg,
+                'input_tokens': 0,
+                'output_tokens': 0,
+                'total_tokens': 0,
+                'model': active_model,
+                'cost': 0.0,
+                'api_id': None,
+            }
+        
         tracker.ai_call(f"Using model: {active_model}")
         
         # Step 3: Auto-enable JSON mode for supported models