diff --git a/web/components/UI.tsx b/web/components/UI.tsx
index 787fd2a..ad9390d 100644
--- a/web/components/UI.tsx
+++ b/web/components/UI.tsx
@@ -219,14 +219,14 @@ export const Drawer: React.FC<DrawerProps> = ({ isOpen, onClose, title, classNam
       <div className="fixed inset-0 bg-black/60 backdrop-blur-sm transition-opacity" onClick={onClose} />
       
       {/* Drawer Content */}
-      <div className={`relative ml-auto flex h-full w-full flex-col overflow-y-auto bg-background/95 border-l border-white/10 p-6 shadow-2xl animate-in slide-in-from-right ${className || 'max-w-md sm:max-w-lg'}`}>
-        <div className="flex items-center justify-between mb-4">
+      <div className={`relative ml-auto flex h-full w-full flex-col bg-background/95 border-l border-white/10 p-6 shadow-2xl animate-in slide-in-from-right ${className || 'max-w-md sm:max-w-lg'}`}>
+        <div className="flex items-center justify-between mb-4 shrink-0">
           <h2 className="text-lg font-semibold text-foreground">{title}</h2>
           <Button variant="ghost" size="icon" onClick={onClose}>
             <X className="h-4 w-4" />
           </Button>
         </div>
-        <div className="flex-1">
+        <div className="flex-1 min-h-0 overflow-y-auto">
           {children}
         </div>
       </div>
diff --git a/web/pages/Assistants.tsx b/web/pages/Assistants.tsx
index e116d17..cee26b1 100644
--- a/web/pages/Assistants.tsx
+++ b/web/pages/Assistants.tsx
@@ -2379,7 +2379,7 @@ export const DebugDrawer: React.FC<{
   const [inputText, setInputText] = useState('');
   const [isLoading, setIsLoading] = useState(false);
   const [callStatus, setCallStatus] = useState<'idle' | 'calling' | 'active'>('idle');
-  const [agentState, setAgentState] = useState<'waiting' | 'listening' | 'thinking' | 'speaking'>('waiting');
+  const [agentState, setAgentState] = useState<'listening' | 'thinking' | 'speaking'>('listening');
   const [textPromptDialog, setTextPromptDialog] = useState<DebugTextPromptDialogState>({
     open: false,
     message: '',
@@ -2562,7 +2562,7 @@ export const DebugDrawer: React.FC<{
         clearResponseTracking();
         setMessages([]);
         setCallStatus('idle');
-        setAgentState('waiting');
+        setAgentState('listening');
       }
     } else {
         setMode('text');
@@ -2586,7 +2586,7 @@ export const DebugDrawer: React.FC<{
         setSettingsDrawerOpen(false);
         setIsSwapped(false);
         setCallStatus('idle');
-        setAgentState('waiting');
+        setAgentState('listening');
     }
   }, [isOpen, assistant, mode]);
 
@@ -3112,7 +3112,7 @@ export const DebugDrawer: React.FC<{
         console.error(e);
         stopVoiceCapture();
         setCallStatus('idle');
-        setAgentState('waiting');
+        setAgentState('listening');
         const err = e as Error & { __dynamicVariables?: boolean };
         if (err.__dynamicVariables) {
           setWsStatus('disconnected');
@@ -3138,7 +3138,7 @@ export const DebugDrawer: React.FC<{
     stopMedia();
     closeWs();
     setCallStatus('idle');
-    setAgentState('waiting');
+    setAgentState('listening');
     clearResponseTracking();
     setMessages([]);
     setTextPromptDialog({ open: false, message: '', promptType: 'text' });
@@ -3590,7 +3590,7 @@ export const DebugDrawer: React.FC<{
         }
 
         if (type === 'output.audio.end') {
-          setAgentState('waiting');
+          setAgentState('listening');
           return;
         }
 
@@ -4516,36 +4516,31 @@ export const DebugDrawer: React.FC<{
                         </div>
                         <div className="text-center z-10">
                             <h3 className="text-xl font-bold text-white mb-2 tracking-tight">
-                              {agentState === 'listening' ? '正在倾听...' :
-                               agentState === 'thinking' ? '思考中...' :
+                              {agentState === 'thinking' ? '思考中...' :
                                agentState === 'speaking' ? '正在回复...' :
-                               '待机中'}
+                               '正在倾听...'}
                             </h3>
                             <div className="flex items-center justify-center gap-2">
                                 <span className="relative flex h-2.5 w-2.5">
                                   <span className={`animate-ping absolute inline-flex h-full w-full rounded-full opacity-75 ${
-                                    agentState === 'listening' ? 'bg-blue-400' :
                                     agentState === 'thinking' ? 'bg-yellow-400' :
                                     agentState === 'speaking' ? 'bg-green-400' :
-                                    'bg-gray-400'
+                                    'bg-blue-400'
                                   }`}></span>
                                   <span className={`relative inline-flex rounded-full h-2.5 w-2.5 ${
-                                    agentState === 'listening' ? 'bg-blue-500' :
                                     agentState === 'thinking' ? 'bg-yellow-500' :
                                     agentState === 'speaking' ? 'bg-green-500' :
-                                    'bg-gray-500'
+                                    'bg-blue-500'
                                   }`}></span>
                                 </span>
                                 <p className={`text-sm font-medium ${
-                                    agentState === 'listening' ? 'text-blue-400' :
                                     agentState === 'thinking' ? 'text-yellow-400' :
                                     agentState === 'speaking' ? 'text-green-400' :
-                                    'text-gray-400'
+                                    'text-blue-400'
                                 }`}>
-                                  {agentState === 'listening' ? 'Listening' :
-                                   agentState === 'thinking' ? 'Thinking' :
+                                  {agentState === 'thinking' ? 'Thinking' :
                                    agentState === 'speaking' ? 'Speaking' :
-                                   'Waiting'}
+                                   'Listening'}
                                 </p>
                             </div>
                         </div>