标题生成强制关闭 think 模式,回答速度指示器默认开启

- TitleMiddleware 给模型实例注入 extra_body 关 think(覆盖 Zhipu GLM 与 Qwen/vLLM 两套写法),并兜底剥离不闭合/残留的 <think> 标签,避免 GLM5 起标题时长时间思考与标签泄漏。
- 前端 streamSpeedMeterEnabled 默认改为 true,设置项描述同步为"默认开启"。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

frontend-web/src/components/workspace/settings/appearance-settings-page.tsx

@@ -122,7 +122,7 @@ export function AppearanceSettingsPage() {
             <div className="text-sm font-medium">回答速度指示器</div>
             <p className="text-muted-foreground mt-1 max-w-md text-xs leading-relaxed">
               在生成回答时显示一个小胶囊，实时估算当前 token/s
-              并用绿到红的渐变条直观反映快慢。默认关闭。
+              并用绿到红的渐变条直观反映快慢。默认开启。
             </p>
           </div>
           <Switch

frontend-web/src/core/settings/local.ts

@@ -5,15 +5,15 @@ export const DEFAULT_LOCAL_SETTINGS: LocalSettings = {
     enabled: true,
   },
   display: {
-    streamSpeedMeterEnabled: false,
+    streamSpeedMeterEnabled: true,
   },
   context: {
-    model_name: undefined,
-    mode: undefined,
-    reasoning_effort: undefined,
-    writing_mode: false,
-  },
-};
+    model_name: undefined,
+    mode: undefined,
+    reasoning_effort: undefined,
+    writing_mode: false,
+  },
+};
 
 export const LOCAL_SETTINGS_KEY = "deerflow.local-settings";
 export const THREAD_MODEL_KEY_PREFIX = "deerflow.thread-model.";
@@ -29,7 +29,7 @@ export interface LocalSettings {
   display: {
     /**
      * Whether to render a small token/s speed meter while a streaming
-     * AI reply is in flight. Default off; controlled by the user from
+     * AI reply is in flight. Default on; controlled by the user from
      * the appearance settings page.
      */
     streamSpeedMeterEnabled: boolean;