fix(ui): show error details in title generation error dialog

Include the exception message in the title generation error dialog, matching
completion, transcription, and save error handling. Change generate_title
to return (title, error) so the presenter can display the real error.

Author: AAClause

basilisk/presenters/conversation_presenter.py

@@ -452,7 +452,7 @@ def generate_conversation_title(self):
 		model = self.view.current_model
 		if not model:
 			return
-		title = self.service.generate_title(
+		title, error = self.service.generate_title(
 			engine=self.view.current_engine,
 			conversation=self.conversation,
 			provider_id=self.view.current_account.provider.id,
@@ -463,10 +463,13 @@ def generate_conversation_title(self):
 			stream=self.view.stream_mode.GetValue(),
 		)
 		if title is None and self.conversation.messages:
+			message = (
+				_("An error occurred during title generation: %s") % error
+				if error
+				else _("An error occurred during title generation")
+			)
 			self.view.show_enhanced_error(
-				_("An error occurred during title generation"),
-				_("Title Generation Error"),
-				is_completion_error=True,
+				message, _("Title Generation Error"), is_completion_error=True
 			)
 		return title
 

basilisk/services/conversation_service.py

@@ -209,7 +209,7 @@ def generate_title(
 		top_p: float,
 		max_tokens: int,
 		stream: bool = True,
-	) -> Optional[str]:
+	) -> tuple[Optional[str], Optional[Exception]]:
 		"""Generate a conversation title using the AI model.
 
 		Streaming is used by default; some providers (e.g. Anthropic) require
@@ -228,7 +228,8 @@ def generate_title(
 				models that do not support streaming.
 
 		Returns:
-			The generated title string, or None on failure.
+			A tuple of (title, error). On success: (title_string, None).
+			On failure: (None, exception).
 		"""
 		play_sound("progress", loop=True)
 		try:
@@ -254,13 +255,13 @@ def generate_title(
 				for chunk in engine.completion_response_with_stream(response):
 					if isinstance(chunk, str):
 						content_parts.append(chunk)
-				return "".join(content_parts).strip()
+				return "".join(content_parts).strip(), None
 			new_block = engine.completion_response_without_stream(
 				response=response, **completion_kw
 			)
-			return new_block.response.content
-		except Exception:
+			return new_block.response.content, None
+		except Exception as e:
 			log.error("Title generation failed", exc_info=True)
-			return None
+			return None, e
 		finally:
 			stop_sound()

tests/presenters/test_conversation_presenter.py

@@ -404,7 +404,7 @@ def test_returns_title_on_success(
 			model=AIModelInfo(provider_id="openai", model_id="test"),
 		)
 		presenter.conversation.add_block(block, None)
-		mock_service.generate_title.return_value = "My Conversation"
+		mock_service.generate_title.return_value = ("My Conversation", None)
 
 		mocker.patch.object(
 			presenter.completion_handler, "is_running", return_value=False
@@ -416,14 +416,19 @@ def test_returns_title_on_success(
 	def test_shows_error_when_title_none_but_messages_exist(
 		self, presenter, mock_view, mock_service, mocker
 	):
-		"""Shows enhanced error dialog when service returns None but messages exist."""
+		"""Shows enhanced error dialog with error details when service fails."""
 		block = MessageBlock(
 			request=Message(role=MessageRoleEnum.USER, content="Hi"),
 			response=Message(role=MessageRoleEnum.ASSISTANT, content="Hello"),
 			model=AIModelInfo(provider_id="openai", model_id="test"),
 		)
 		presenter.conversation.add_block(block, None)
-		mock_service.generate_title.return_value = None
+		mock_service.generate_title.return_value = (
+			None,
+			ValueError(
+				"Streaming is required for operations that may take longer than 10 minutes."
+			),
+		)
 
 		mocker.patch.object(
 			presenter.completion_handler, "is_running", return_value=False
@@ -432,3 +437,5 @@ def test_shows_error_when_title_none_but_messages_exist(
 
 		assert result is None
 		mock_view.show_enhanced_error.assert_called_once()
+		call_args = mock_view.show_enhanced_error.call_args
+		assert "Streaming is required" in call_args[0][0]

tests/services/test_conversation_service.py

@@ -189,7 +189,7 @@ def test_returns_content(self, service):
 				model=AIModelInfo(provider_id="openai", model_id="test"),
 			)
 		)
-		title = service.generate_title(
+		title, error = service.generate_title(
 			engine=mock_engine,
 			conversation=conv,
 			provider_id="openai",
@@ -199,6 +199,7 @@ def test_returns_content(self, service):
 			max_tokens=100,
 		)
 		assert title == "My Title"
+		assert error is None
 
 	def test_returns_content_without_stream(self, service):
 		"""generate_title with stream=False should use non-streaming path."""
@@ -221,7 +222,7 @@ def test_returns_content_without_stream(self, service):
 				model=AIModelInfo(provider_id="openai", model_id="test"),
 			)
 		)
-		title = service.generate_title(
+		title, error = service.generate_title(
 			engine=mock_engine,
 			conversation=conv,
 			provider_id="openai",
@@ -232,14 +233,15 @@ def test_returns_content_without_stream(self, service):
 			stream=False,
 		)
 		assert title == "Non-Stream Title"
+		assert error is None
 
 	def test_returns_none_on_error(self, service):
-		"""generate_title should return None on exception."""
+		"""generate_title should return (None, exception) on exception."""
 		mock_engine = MagicMock()
 		mock_engine.completion.side_effect = RuntimeError("API down")
 
 		conv = Conversation()
-		title = service.generate_title(
+		title, error = service.generate_title(
 			engine=mock_engine,
 			conversation=conv,
 			provider_id="openai",
@@ -249,3 +251,5 @@ def test_returns_none_on_error(self, service):
 			max_tokens=100,
 		)
 		assert title is None
+		assert error is not None
+		assert str(error) == "API down"