feat: LLM_QUESTION_MARK_REQUIRED added as an optional gate for AI answer generation

README.md

@@ -44,6 +44,7 @@ Configure via the environment variables:
 - `LLM_CONTEXT_SHALLOW_COUNT`: Results with headlines only (additional breadth). Default `15`.
 - `LLM_TABS`: Tab whitelist, comma delimiter. Default `general,science,it,news`.
 - `LLM_INTERACTIVE`: UI mode. Default is `true` (interactive: copy, regenerate, follow up). Set to `false` for simple response only mode.
+- `LLM_QUESTION_MARK_REQUIRED`: Only trigger AI answers when the query contains `?`. Default `false`.
 
 ## How It Works
 1 user initial search 

ai_answers.py

@@ -440,6 +440,7 @@ class SXNGPlugin(Plugin):
 
     def _load_config(self):
         self.interactive = os.getenv('LLM_INTERACTIVE', 'true').lower().strip() in ('true', '1', 'yes', 'on')
+        self.question_mark_required = os.getenv('LLM_QUESTION_MARK_REQUIRED', 'false').lower().strip() in ('true', '1', 'yes', 'on')
         raw_provider = os.getenv('LLM_PROVIDER', '').lower().strip()
         
         raw_url = os.getenv('LLM_URL', '').strip()
@@ -899,6 +900,9 @@ class SXNGPlugin(Plugin):
             if request and hasattr(request, 'headers') and request.headers.get('X-AI-Auxiliary'):
                 return results
 
+            if self.question_mark_required and '?' not in search.search_query.query:
+                return results
+
             current_tabs = set(search.search_query.categories)
             if not current_tabs: current_tabs = {'general'}