fix: implement real-time streaming for responses

- Added streaming service support to the Q&A agent for real-time token streaming.
- Updated `answer_question` method to stream responses token-by-token to the frontend.
- Modified `handle_qna_workflow` to handle both custom and values streaming modes.
- Enhanced state management to include streaming service for improved user experience.

surfsense_web/app/dashboard/[search_space_id]/researcher/[[...chat_id]]/page.tsx

@@ -17,13 +17,14 @@ export default function ResearcherPage() {
 	const { search_space_id } = useParams();
 	const router = useRouter();
 	const hasSetInitialConnectors = useRef(false);
+	const hasInitiatedResponse = useRef<string | null>(null);
 	const activeChatId = useAtomValue(activeChatIdAtom);
 	const { data: activeChatState, isFetching: isChatLoading } = useAtomValue(activeChatAtom);
 	const { mutateAsync: createChat } = useAtomValue(createChatMutationAtom);
 	const { mutateAsync: updateChat } = useAtomValue(updateChatMutationAtom);
 	const isNewChat = !activeChatId;
 
-	// Reset the flag when chat ID changes
+	// Reset the flag when chat ID changes (but not hasInitiatedResponse - we need to remember if we already initiated)
 	useEffect(() => {
 		hasSetInitialConnectors.current = false;
 	}, [activeChatId]);
@@ -167,10 +168,14 @@ export default function ResearcherPage() {
 			if (chatData.messages && Array.isArray(chatData.messages)) {
 				if (chatData.messages.length === 1 && chatData.messages[0].role === "user") {
 					// Single user message - append to trigger LLM response
-					handler.append({
-						role: "user",
-						content: chatData.messages[0].content,
-					});
+					// Only if we haven't already initiated for this chat and handler doesn't have messages yet
+					if (hasInitiatedResponse.current !== activeChatId && handler.messages.length === 0) {
+						hasInitiatedResponse.current = activeChatId;
+						handler.append({
+							role: "user",
+							content: chatData.messages[0].content,
+						});
+					}
 				} else if (chatData.messages.length > 1) {
 					// Multiple messages - set them all
 					handler.setMessages(chatData.messages);