Fix voice command data structure and prevent duplicate activities

Backend changes:
- Update LLM prompt to use correct field names matching frontend interfaces
  - Use 'diaperType' instead of 'type' for diaper activities
  - Use 'feedingType' instead of 'method' for feeding activities
  - Simplify sleep structure (duration, quality, location only)

Frontend changes:
- Add processedClassificationId tracking to prevent infinite loop
- Create unique ID for each classification to avoid duplicate processing
- Reset processed ID when dialog opens/closes or new recording starts

This fixes the issue where voice commands created multiple duplicate
activities and had mismatched data structures causing tracker warnings.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

maternal-web/components/voice/VoiceFloatingButton.tsx

@@ -39,6 +39,7 @@ export function VoiceFloatingButton() {
   const [processingStatus, setProcessingStatus] = useState<'listening' | 'understanding' | 'saving' | null>(null);
   const [identifiedActivity, setIdentifiedActivity] = useState<string>('');
   const [classificationResult, setClassificationResult] = useState<any>(null);
+  const [processedClassificationId, setProcessedClassificationId] = useState<string | null>(null);
   const [snackbar, setSnackbar] = useState<{
     open: boolean;
     message: string;
@@ -67,11 +68,19 @@ export function VoiceFloatingButton() {
   // Auto-use classification from backend when transcription completes
   // MediaRecorder sends audio to backend, which transcribes + classifies in one call
   React.useEffect(() => {
-    if (classification && !isListening && !isProcessing && open) {
+    // Create a unique ID for this classification based on transcript + type + timestamp
+    const classificationId = classification
+      ? `${transcript}-${classification.type}-${classification.timestamp}`
+      : null;
+
+    // Only process if we haven't already processed this exact classification
+    if (classification && !isListening && !isProcessing && open && classificationId !== processedClassificationId) {
+      console.log('[Voice] New classification detected, processing...', classificationId);
+      setProcessedClassificationId(classificationId);
       setClassificationResult(classification);
       handleClassifiedIntent(classification);
     }
-  }, [classification, isListening, isProcessing, open]);
+  }, [classification, isListening, isProcessing, open, transcript, processedClassificationId]);
 
   const handleOpen = () => {
     if (!isSupported) {
@@ -87,6 +96,7 @@ export function VoiceFloatingButton() {
     setClassificationResult(null);
     setProcessingStatus(null);
     setIdentifiedActivity('');
+    setProcessedClassificationId(null);
   };
 
   const handleClose = () => {
@@ -98,11 +108,13 @@ export function VoiceFloatingButton() {
     setClassificationResult(null);
     setProcessingStatus(null);
     setIdentifiedActivity('');
+    setProcessedClassificationId(null);
   };
 
   const handleStartListening = () => {
     reset();
     setClassificationResult(null);
+    setProcessedClassificationId(null);
     startListening();
   };