@ww_nero/media 1.1.0 → 1.2.0

package/index.js

@@ -114,7 +114,7 @@ const resolveAudioFile = (workingDir, rawPath) => {
 const server = new Server(
   {
     name: 'media',
-    version: '1.1.0',
+    version: '1.2.0',
   },
   {
     capabilities: {
@@ -157,6 +157,11 @@ server.setRequestHandler(ListToolsRequestSchema, async () => ({
             type: 'string',
             description: '需要合成语音的文本内容',
           },
+          gender: {
+            type: 'string',
+            enum: ['female', 'male'],
+            description: '语音性别，female 为女声，male 为男声，默认为 female',
+          },
         },
         required: ['working_directory', 'text'],
       },
@@ -180,7 +185,7 @@ server.setRequestHandler(CallToolRequestSchema, async (request) => {
     }
 
     if (name === 'tts') {
-      const { working_directory, text } = args;
+      const { working_directory, text, gender } = args;
       if (!working_directory || !text) {
         throw new Error('必须同时提供 working_directory 和 text 参数');
       }
@@ -188,6 +193,7 @@ server.setRequestHandler(CallToolRequestSchema, async (request) => {
       const filename = await tts({
         workingDir,
         text,
+        gender,
         apiKey: DASHSCOPE_API_KEY,
       });
       return { content: [{ type: 'text', text: `语音合成完成，音频文件已保存到工作目录下：${filename}` }] };

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@ww_nero/media",
-  "version": "1.1.0",
+  "version": "1.2.0",
   "description": "MCP server for media processing, including ASR speech recognition and TTS speech synthesis",
   "main": "index.js",
   "bin": {

package/utils/tts.js

@@ -5,12 +5,18 @@ const { v4: uuidv4 } = require('uuid');
 
 const TTS_WS_URL = 'wss://dashscope.aliyuncs.com/api-ws/v1/inference/';
 
+// 音色配置
+const VOICE_MAP = {
+  female: 'sambert-zhimiao-emo-v1',
+  male: 'sambert-zhihao-v1',
+};
+
 /**
  * TTS 语音合成
  */
-const tts = async ({ workingDir, text, apiKey }) => {
-  // 固定参数
-  const voice = 'sambert-zhimiao-emo-v1';
+const tts = async ({ workingDir, text, gender = 'female', apiKey }) => {
+  // 根据性别选择音色
+  const voice = VOICE_MAP[gender] || VOICE_MAP.female;
   const format = 'mp3';
   const sampleRate = 16000;