Add tts/text output schema

web/pages/Assistants.tsx

@@ -118,6 +118,7 @@ export const AssistantsPage: React.FC = () => {
       prompt: '',
       knowledgeBaseId: '',
       language: 'zh',
+      voiceOutputEnabled: true,
       voice: voices[0]?.id || '',
       speed: 1,
       hotwords: [],
@@ -531,6 +532,7 @@ export const AssistantsPage: React.FC = () => {
                                         placeholder="设定小助手的人设、语气、行为规范以及业务逻辑..."
                                     />
                                 </div>
+
                             </div>
                         )}
 
@@ -624,15 +626,32 @@ export const AssistantsPage: React.FC = () => {
                                     </p>
                                 </div>
 
+                                <div className="space-y-2">
+                                    <label className="text-sm font-medium text-white flex items-center">
+                                        <Volume2 className="w-4 h-4 mr-2 text-primary" /> 启用语音输出
+                                    </label>
+                                    <label className="flex h-12 items-center justify-between rounded-xl border border-white/10 bg-white/5 px-4 text-sm">
+                                        <span className="text-foreground">TTS 输出</span>
+                                        <input
+                                          type="checkbox"
+                                          checked={selectedAssistant.voiceOutputEnabled !== false}
+                                          onChange={(e) => updateAssistant('voiceOutputEnabled', e.target.checked)}
+                                          className="accent-primary"
+                                        />
+                                    </label>
+                                    <p className="text-xs text-muted-foreground">关闭后将进入纯文本输出模式，不会产生语音音频。</p>
+                                </div>
+
                                 <div className="space-y-2">
                                     <label className="text-sm font-medium text-white flex items-center">
                                         <Volume2 className="w-4 h-4 mr-2 text-primary"/> 选择音色 (From Voice Library)
                                     </label>
                                     <div className="relative group">
                                         <select 
-                                            className="flex h-12 w-full rounded-xl border border-white/10 bg-white/5 px-4 py-1 text-sm shadow-sm transition-all focus-visible:outline-none focus-visible:ring-1 focus-visible:ring-primary/50 [&>option]:bg-card text-foreground appearance-none cursor-pointer"
+                                            className="flex h-12 w-full rounded-xl border border-white/10 bg-white/5 px-4 py-1 text-sm shadow-sm transition-all focus-visible:outline-none focus-visible:ring-1 focus-visible:ring-primary/50 [&>option]:bg-card text-foreground appearance-none cursor-pointer disabled:opacity-50 disabled:cursor-not-allowed"
                                             value={selectedAssistant.voice}
                                             onChange={(e) => updateAssistant('voice', e.target.value)}
+                                            disabled={selectedAssistant.voiceOutputEnabled === false}
                                         >
                                             <option value="" disabled>请选择声音库中的声音...</option>
                                             {voices.map(voice => (
@@ -645,7 +664,9 @@ export const AssistantsPage: React.FC = () => {
                                     </div>
                                     <p className="text-xs text-muted-foreground flex items-center mt-1">
                                         <Sparkles className="w-3 h-3 mr-1 text-primary opacity-70" />
-                                        音色配置同步自声音库。如需添加更多音色，请前往“声音库”模块。
+                                        {selectedAssistant.voiceOutputEnabled === false
+                                          ? '启用语音输出后才可选择音色。'
+                                          : '音色配置同步自声音库。如需添加更多音色，请前往“声音库”模块。'}
                                     </p>
                                 </div>
 
@@ -916,6 +937,7 @@ export const AssistantsPage: React.FC = () => {
            llmModels={llmModels}
            asrModels={asrModels}
            tools={tools}
+           textTtsEnabled={selectedAssistant.voiceOutputEnabled !== false}
          />
       )}
 
@@ -1030,6 +1052,7 @@ export const DebugDrawer: React.FC<{
   llmModels?: LLMModel[];
   asrModels?: ASRModel[];
   tools?: Tool[];
+  textTtsEnabled: boolean;
   sessionMetadataExtras?: Record<string, any>;
   onProtocolEvent?: (event: Record<string, any>) => void;
 }> = ({
@@ -1040,6 +1063,7 @@ export const DebugDrawer: React.FC<{
   llmModels = [],
   asrModels = [],
   tools = [],
+  textTtsEnabled,
   sessionMetadataExtras,
   onProtocolEvent,
 }) => {
@@ -1117,7 +1141,6 @@ export const DebugDrawer: React.FC<{
   const [selectedCamera, setSelectedCamera] = useState<string>('');
   const [selectedMic, setSelectedMic] = useState<string>('');
   const [isSwapped, setIsSwapped] = useState(false); 
-  const [textTtsEnabled, setTextTtsEnabled] = useState(true);
   const [aecEnabled, setAecEnabled] = useState<boolean>(() => localStorage.getItem('debug_audio_aec') !== '0');
   const [nsEnabled, setNsEnabled] = useState<boolean>(() => localStorage.getItem('debug_audio_ns') !== '0');
   const [agcEnabled, setAgcEnabled] = useState<boolean>(() => localStorage.getItem('debug_audio_agc') !== '0');
@@ -1546,6 +1569,7 @@ export const DebugDrawer: React.FC<{
   const buildLocalResolvedRuntime = () => {
     const warnings: string[] = [];
     const services: Record<string, any> = {};
+    const ttsEnabled = Boolean(textTtsEnabled);
     const isExternalLlm = assistant.configMode === 'dify' || assistant.configMode === 'fastgpt';
     const knowledgeBaseId = String(assistant.knowledgeBaseId || '').trim();
     const knowledge = knowledgeBaseId
@@ -1597,6 +1621,7 @@ export const DebugDrawer: React.FC<{
       if (voice) {
         const ttsProvider = isSiliconflowVendor(voice.vendor) ? 'siliconflow' : 'edge';
         services.tts = {
+          enabled: ttsEnabled,
           provider: ttsProvider,
           model: voice.model,
           apiKey: ttsProvider === 'siliconflow' ? voice.apiKey : null,
@@ -1605,17 +1630,25 @@ export const DebugDrawer: React.FC<{
         };
       } else {
         services.tts = {
+          enabled: ttsEnabled,
           voice: assistant.voice,
           speed: assistant.speed || 1.0,
         };
         warnings.push(`Voice resource not found in loaded list: ${assistant.voice}`);
       }
+    } else if (!ttsEnabled) {
+      services.tts = {
+        enabled: false,
+      };
     }
 
     const localResolved = {
       assistantId: assistant.id,
       warnings,
       sessionStartMetadata: {
+        output: {
+          mode: ttsEnabled ? 'audio' : 'text',
+        },
         systemPrompt: assistant.prompt || '',
         greeting: assistant.opener || '',
         knowledgeBaseId,
@@ -2007,15 +2040,9 @@ export const DebugDrawer: React.FC<{
         </div>
         <div className="flex items-center justify-between gap-2">
           <Badge variant="outline" className="text-xs">WS: {wsStatus}</Badge>
-          <label className="inline-flex items-center gap-1 text-xs text-muted-foreground px-2 py-1 rounded border border-white/10">
-            <input
-              type="checkbox"
-              checked={textTtsEnabled}
-              onChange={(e) => setTextTtsEnabled(e.target.checked)}
-              className="accent-primary"
-            />
-            TTS
-          </label>
+          <Badge variant={textTtsEnabled ? 'outline' : 'secondary'} className="text-xs">
+            TTS: {textTtsEnabled ? 'ON' : 'OFF'}
+          </Badge>
         </div>
         <div className="rounded-md border border-white/10 bg-black/20 p-2 space-y-2">
           <p className="text-[10px] uppercase tracking-widest text-muted-foreground">Audio 3A</p>

web/services/backendApi.ts

@@ -33,6 +33,7 @@ const mapAssistant = (raw: AnyRecord): Assistant => ({
   prompt: readField(raw, ['prompt'], ''),
   knowledgeBaseId: readField(raw, ['knowledgeBaseId', 'knowledge_base_id'], ''),
   language: readField(raw, ['language'], 'zh') as 'zh' | 'en',
+  voiceOutputEnabled: Boolean(readField(raw, ['voiceOutputEnabled', 'voice_output_enabled'], true)),
   voice: readField(raw, ['voice'], ''),
   speed: Number(readField(raw, ['speed'], 1)),
   hotwords: readField(raw, ['hotwords'], []),
@@ -210,6 +211,7 @@ export const createAssistant = async (data: Partial<Assistant>): Promise<Assista
     prompt: data.prompt || '',
     knowledgeBaseId: data.knowledgeBaseId || '',
     language: data.language || 'zh',
+    voiceOutputEnabled: data.voiceOutputEnabled ?? true,
     voice: data.voice || '',
     speed: data.speed ?? 1,
     hotwords: data.hotwords || [],
@@ -234,6 +236,7 @@ export const updateAssistant = async (id: string, data: Partial<Assistant>): Pro
     prompt: data.prompt,
     knowledgeBaseId: data.knowledgeBaseId,
     language: data.language,
+    voiceOutputEnabled: data.voiceOutputEnabled,
     voice: data.voice,
     speed: data.speed,
     hotwords: data.hotwords,

web/types.ts

@@ -7,6 +7,7 @@ export interface Assistant {
   prompt: string;
   knowledgeBaseId: string;
   language: 'zh' | 'en';
+  voiceOutputEnabled?: boolean;
   voice: string; // This will now store the ID of the voice from Voice Library
   speed: number;
   hotwords: string[];