myagent-ai 1.16.3 → 1.16.4

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "myagent-ai",
-  "version": "1.16.3",
+  "version": "1.16.4",
   "description": "本地桌面端执行型AI助手 - Open Interpreter 风格 | Local Desktop Execution-Oriented AI Assistant",
   "main": "main.py",
   "bin": {

package/web/api_server.py

@@ -1093,10 +1093,19 @@ class ApiServer:
 
             # ── 尝试 faster-whisper ──
             try:
+                # [v1.16.3] 抑制 HuggingFace 未认证警告和 pydub ffmpeg 警告
+                import warnings
+                warnings.filterwarnings("ignore", message=".*HF_TOKEN.*", category=UserWarning)
+                warnings.filterwarnings("ignore", message=".*huggingface_hub.*token.*", category=UserWarning)
+                warnings.filterwarnings("ignore", message=".*ffmpeg or avconv.*", category=RuntimeWarning)
+                # 抑制 HF Hub telemetry 和进度条
+                os.environ.setdefault("HF_HUB_DISABLE_TELEMETRY", "1")
+                os.environ.setdefault("HF_HUB_DISABLE_PROGRESS_BARS", "1")
+                os.environ.setdefault("TRANSFORMERS_VERBOSITY", "error")
+
                 from faster_whisper import WhisperModel
                 whisper_model = getattr(self, '_whisper_model', None)
                 if whisper_model is None:
-                    import os
                     model_dir = os.path.join(os.path.dirname(os.path.dirname(os.path.abspath(__file__))), 'models', 'whisper')
                     # 使用 tiny 模型（最轻量，~39MB），CPU int8 量化
                     self._whisper_model = WhisperModel("tiny", device="cpu", compute_type="int8",

package/web/ui/chat/chat.css

@@ -720,9 +720,21 @@ input,textarea,select{font:inherit}
   line-height: 1.6;
   white-space: pre-wrap;
   word-break: break-word;
-  min-height: 24px;
+  min-height: 48px;
   max-height: 120px;
-  overflow-y: auto;
+  width: 100%;
+  border: 1px solid var(--border);
+  border-radius: var(--radius-sm);
+  padding: 8px 10px;
+  background: var(--bg);
+  resize: none;
+  outline: none;
+  font-family: inherit;
+  box-sizing: border-box;
+  transition: border-color 0.15s;
+}
+.voice-preview-text:focus {
+  border-color: var(--accent);
 }
 .voice-preview-actions {
   display: flex;

package/web/ui/chat/chat_container.html

@@ -177,7 +177,7 @@
           </div>
           <div class="voice-preview" id="voicePreview" style="display:none">
             <div class="voice-preview-label">语音输入 · <span id="voicePreviewHint">识别中...</span></div>
-            <div class="voice-preview-text" id="voicePreviewText"></div>
+            <textarea class="voice-preview-text" id="voicePreviewText" placeholder="识别结果..." rows="2"></textarea>
             <div class="voice-preview-actions">
               <button class="voice-preview-cancel" onclick="cancelVoicePreview()">取消</button>
               <button class="voice-preview-send" id="voicePreviewSend" onclick="sendVoiceMessage()">发送</button>

package/web/ui/chat/chat_main.js

@@ -4590,7 +4590,7 @@ var VoiceInput = {
     // 显示预览区域
     if (voiceArea) voiceArea.style.display = 'none';
     if (voicePreview) voicePreview.style.display = 'block';
-    if (previewText) previewText.textContent = '识别中...';
+    if (previewText) previewText.value = '识别中...';
     if (previewHint) previewHint.textContent = '正在发送到本地STT引擎';
     if (previewSend) previewSend.disabled = true;
 
@@ -4614,11 +4614,11 @@ var VoiceInput = {
       if (data && data.text && data.text.trim()) {
         this.rawText = data.text.trim();
         this._sttEngine = data.engine || 'unknown';
-        if (previewText) previewText.textContent = this.rawText;
-        if (previewHint) previewHint.textContent = '已识别 (' + (this._sttEngine || 'local') + ')';
+        if (previewText) previewText.value = this.rawText;
+        if (previewHint) previewHint.textContent = '已识别 (' + (this._sttEngine || 'local') + ')，可编辑后发送';
       } else if (data && data.error) {
         if (previewHint) previewHint.textContent = '识别失败';
-        if (previewText) previewText.textContent = data.error;
+        if (previewText) previewText.value = data.error;
         console.warn('Voice STT error:', data.error);
         // 如果没有STT引擎，给出提示
         if (data.available === false) {
@@ -4629,12 +4629,12 @@ var VoiceInput = {
         }
       } else {
         if (previewHint) previewHint.textContent = '未识别到文字';
-        if (previewText) previewText.textContent = '未识别到文字，请重试';
+        if (previewText) previewText.value = '未识别到文字，请重试';
       }
     } catch (e) {
       console.error('Voice STT request error:', e);
       if (previewHint) previewHint.textContent = '网络错误';
-      if (previewText) previewText.textContent = 'STT请求失败，请重试';
+      if (previewText) previewText.value = 'STT请求失败，请重试';
     }
 
     this._cleanupStream();
@@ -4669,11 +4669,12 @@ var VoiceInput = {
     this.audioChunks = [];
   },
 
-  /** Send the voice text as a message（直接发送原始STT文本，由后端usersays_correct纠错） */
+  /** Send the voice text as a message（支持用户编辑识别结果后再发送） */
   sendMessage: function() {
-    if (!this.rawText || !this.rawText.trim()) return;
-
-    var text = this.rawText.trim();
+    // 从 textarea 读取编辑后的文本
+    var previewText = document.getElementById('voicePreviewText');
+    var text = previewText ? previewText.value.trim() : '';
+    if (!text) return;
 
     // 切回文本模式
     this.switchMode('text');