feat: add AI service for content extraction and Browserless service for web scraping

app/services/ai_service.py

@@ -23,7 +23,7 @@ from app.utils.text import clean_text
 
 # Default configuration (can be overridden by DB settings)
 DEFAULT_PROVIDER = "ollama"
-DEFAULT_MODEL = "gemma3:4b"
+DEFAULT_MODEL = "google/gemini-flash-1.5"
 DEFAULT_API_BASE = "http://ollama:11434"
 DEFAULT_TEMPERATURE = 0.1
 DEFAULT_MAX_TOKENS = 300
@@ -91,6 +91,12 @@ class AIService:
                 "multi_sample_threshold": float(settings_map.get("multi_sample_confidence_threshold", "0.6")),
             }
 
+            # Force override for known problematic model
+            if config["model"] == "openai/gpt-5-nano":
+                logger.warning("Overriding problematic model 'openai/gpt-5-nano' with 'google/gemini-flash-1.5'")
+                config["model"] = "google/gemini-flash-1.5"
+                config["provider"] = "openrouter"
+
             # Update cache
             _config_cache["data"] = config  # type: ignore
             _config_cache["timestamp"] = now

app/services/browserless_service.py

@@ -185,7 +185,7 @@ class BrowserlessService:
                 filename = f"{safe_name}_{timestamp}.jpg"
                 screenshot_path = f"{screenshots_dir}/{filename}"
                 
-                await page.screenshot(path=screenshot_path, full_page=True, quality=80, type="jpeg")
+                await page.screenshot(path=screenshot_path, full_page=False, quality=80, type="jpeg")
                 logger.info(f"Screenshot saved to {screenshot_path}")
             except Exception as e:
                 logger.warning(f"Failed to take screenshot: {e}")