♻️ refactor(help): 优化帮助页面内容和i18n

- 重新组织帮助页面内容结构
- 优化中英文翻译键的命名
- 改进内容展示的可读性

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: dtsola

frontend/src/renderer/src/locales/en-US.ts

@@ -659,7 +659,7 @@ export default {
       },
       semantic: {
         title: 'Semantic Understanding',
-        description: 'BGE-M3 vector embedding + Ollama/OpenAI compatible LLM'
+        description: 'BGE-M3/Cloud embedding vector + Ollama/Cloud LLM understanding'
       }
     },
     tutorial: {
@@ -670,9 +670,9 @@ export default {
           step1: { label: 'Speech Recognition Model:', text: 'Settings → Speech Settings → Select FastWhisper model version (local)' },
           step2: { label: 'Large Language Model:', text: 'Settings → LLM Settings → Choose local Ollama or OpenAI compatible cloud service' },
           step3: { label: 'Vision Understanding Model:', text: 'Settings → Vision Model → Select CN-CLIP model (local)' },
-          step4: { label: 'Text Embedding Model:', text: 'Settings → Embedding Model → Select BGE-M3 model (local)' },
-          step5: { label: 'Cloud Service Config:', text: 'To use cloud LLM, configure API key and endpoint' },
-          tip: 'Tip: Local models offer better privacy, cloud models provide stronger understanding. Choose based on your needs'
+          step4: { label: 'Text Embedding Model:', text: 'Settings → Embedding Model → Choose local BGE-M3 or cloud embedding API' },
+          step5: { label: 'Cloud Service Config:', text: 'To use cloud LLM or cloud embedding, configure API key and endpoint' },
+          tip: 'Tip: Local models offer better privacy, cloud models provide stronger understanding. You choose the balance between performance and privacy'
         }
       },
       panel2: {
@@ -725,14 +725,16 @@ export default {
         header: 'What\'s the difference between local and cloud models?',
         content: {
           comparisonTitle: 'Comparison:',
-          comparison1: { label: 'Local Models (Ollama):', text: 'Run completely offline, data never leaves device, best privacy, requires higher hardware specs' },
+          comparison1: { label: 'Local Models (Ollama/BGE-M3):', text: 'Run completely offline, data never leaves device, best privacy, requires higher hardware specs' },
           comparison2: { label: 'Cloud Models (OpenAI Compatible):', text: 'Search queries sent to cloud, better understanding, no local hardware resources needed' },
+          comparison3: { label: 'Cloud Embedding Models:', text: 'Provide higher quality vector representations for improved search accuracy, queries sent to cloud' },
           choiceTitle: 'How to Choose:',
-          choice1: 'High privacy requirements → Use local Ollama models',
+          choice1: 'High privacy requirements → Use local models (Ollama + BGE-M3)',
           choice2: 'Pursue better understanding → Use cloud LLMs',
-          choice3: 'Limited hardware config → Use cloud LLMs',
-          choice4: 'Need offline use → Use local Ollama models',
-          note: 'Note: Text embedding, speech recognition, image understanding models currently only support local, no data uploaded'
+          choice3: 'Pursue more accurate search → Use cloud embedding models',
+          choice4: 'Limited hardware config → Use cloud services',
+          choice5: 'Need offline use → Use local models',
+          note: 'Note: Local files and index data are always stored locally, never uploaded. Only search queries are sent to cloud (when using cloud models)'
         }
       },
       panel1: {
@@ -762,12 +764,17 @@ export default {
           ollamaCommand1: 'Install model:',
           ollamaCommand2: 'View models:',
           ollamaCommand3: 'Run model:',
-          cloudTitle: 'Cloud OpenAI Compatible Service:',
-          cloud1: 'Check if API key is properly configured',
-          cloud2: 'Verify endpoint URL is accessible',
-          cloud3: 'Check network connection and proxy settings',
-          cloud4: 'Confirm API service is normal (check provider announcements)',
-          cloud5: 'Check account balance (some services charge by usage)'
+          cloudLlmTitle: 'Cloud LLM Service (OpenAI Compatible):',
+          cloudLlm1: 'Check if API key is properly configured',
+          cloudLlm2: 'Verify endpoint URL is accessible',
+          cloudLlm3: 'Check network connection and proxy settings',
+          cloudLlm4: 'Confirm API service is normal (check provider announcements)',
+          cloudLlm5: 'Check account balance (some services charge by usage)',
+          cloudEmbeddingTitle: 'Cloud Embedding Model Service:',
+          cloudEmbedding1: 'Check if API key is properly configured',
+          cloudEmbedding2: 'Verify endpoint URL is accessible',
+          cloudEmbedding3: 'Check network connection and proxy settings',
+          cloudEmbedding4: 'Confirm model name matches what the provider supports'
         }
       },
       panel3: {
@@ -818,7 +825,7 @@ export default {
     },
     about: {
       appDescription: 'Cross-platform local desktop application (Windows/MacOS/Linux) supporting multimodal AI intelligent search. Provides smarter file retrieval experience for knowledge workers through voice, text, and image inputs. Supports both local Ollama and OpenAI-compatible cloud LLMs for flexible deployment.',
-      tagline: 'Local First · Cloud Optional · Privacy Secure',
+      tagline: 'Local First · Cloud Optional · Privacy You Control',
       features: {
         voice: {
           title: 'Voice Search',
@@ -830,12 +837,12 @@ export default {
         },
         semantic: {
           title: 'Semantic Search',
-          description: 'BGE-M3 vector embedding + Ollama/Cloud LLM understanding'
+          description: 'BGE-M3/Cloud embedding vector + Ollama/Cloud LLM understanding'
         }
       },
       techHighlight: {
         title: 'Local First, Cloud Optional',
-        description: 'Supports both local Ollama and OpenAI-compatible cloud LLMs, flexible switching. Local files and index data always stored locally, only search queries sent to cloud (when using cloud models).'
+        description: 'Supports local Ollama/BGE-M3 and cloud OpenAI-compatible APIs, flexible switching. Local files and index data always stored locally, only search queries sent to cloud (when using cloud models). You choose the balance between performance and privacy.'
       }
     }
   }

frontend/src/renderer/src/locales/zh-CN.ts

@@ -665,7 +665,7 @@ export default {
       },
       semantic: {
         title: '语义理解',
-        description: 'BGE-M3向量嵌入 + Ollama/OpenAI兼容大模型'
+        description: 'BGE-M3/云端嵌入向量 + Ollama/云端大模型理解'
       }
     },
     tutorial: {
@@ -676,9 +676,9 @@ export default {
           step1: { label: '语音识别模型：', text: '设置 → 语音设置 → 选择FastWhisper模型版本（本地）' },
           step2: { label: '大语言模型：', text: '设置 → 大语言模型 → 选择本地Ollama或OpenAI兼容云端服务' },
           step3: { label: '视觉理解模型：', text: '设置 → 视觉模型 → 选择CN-CLIP模型（本地）' },
-          step4: { label: '文本嵌入模型：', text: '设置 → 嵌入模型 → 选择BGE-M3模型（本地）' },
-          step5: { label: '云端服务配置：', text: '如需使用云端大模型，配置API密钥和端点地址' },
-          tip: '建议：本地模型隐私更好，云端模型理解能力更强。可根据需求灵活选择'
+          step4: { label: '文本嵌入模型：', text: '设置 → 嵌入模型 → 选择本地BGE-M3或云端嵌入API' },
+          step5: { label: '云端服务配置：', text: '如需使用云端大模型或云端嵌入，配置API密钥和端点地址' },
+          tip: '建议：本地模型隐私更好，云端模型理解能力更强。性能与隐私的权衡由您选择'
         }
       },
       panel2: {
@@ -731,14 +731,16 @@ export default {
         header: '本地模型和云端模型有什么区别？',
         content: {
           comparisonTitle: '对比说明：',
-          comparison1: { label: '本地模型（Ollama）：', text: '完全离线运行，数据不离开设备，隐私性最佳，需要较高硬件配置' },
+          comparison1: { label: '本地模型（Ollama/BGE-M3）：', text: '完全离线运行，数据不离开设备，隐私性最佳，需要较高硬件配置' },
           comparison2: { label: '云端模型（OpenAI兼容）：', text: '搜索查询会发送到云端，理解能力更强，无需本地硬件资源' },
+          comparison3: { label: '云端嵌入模型：', text: '提供更高质量的向量表示，提升搜索准确度，查询会发送到云端' },
           choiceTitle: '如何选择：',
-          choice1: '对隐私要求高 → 使用本地Ollama模型',
+          choice1: '对隐私要求高 → 使用本地模型（Ollama + BGE-M3）',
           choice2: '追求更好的理解能力 → 使用云端大模型',
-          choice3: '硬件配置有限 → 使用云端大模型',
-          choice4: '需要离线使用 → 使用本地Ollama模型',
-          note: '注意：文本嵌入、语音识别、图像理解模型目前仅支持本地运行，不会上传数据'
+          choice3: '追求更精准的搜索 → 使用云端嵌入模型',
+          choice4: '硬件配置有限 → 使用云端服务',
+          choice5: '需要离线使用 → 使用本地模型',
+          note: '注意：本地文件和索引数据始终存储在本地，不上传。仅搜索查询会发送到云端（使用云端模型时）'
         }
       },
       panel1: {
@@ -768,12 +770,17 @@ export default {
           ollamaCommand1: '安装模型：',
           ollamaCommand2: '查看模型：',
           ollamaCommand3: '运行模型：',
-          cloudTitle: '云端OpenAI兼容服务：',
-          cloud1: '检查API密钥是否正确配置',
-          cloud2: '验证端点地址是否可访问',
-          cloud3: '检查网络连接和代理设置',
-          cloud4: '确认API服务是否正常（查看服务商公告）',
-          cloud5: '检查账户余额是否充足（部分服务按量计费）'
+          cloudLlmTitle: '云端大模型服务（OpenAI兼容）：',
+          cloudLlm1: '检查API密钥是否正确配置',
+          cloudLlm2: '验证端点地址是否可访问',
+          cloudLlm3: '检查网络连接和代理设置',
+          cloudLlm4: '确认API服务是否正常（查看服务商公告）',
+          cloudLlm5: '检查账户余额是否充足（部分服务按量计费）',
+          cloudEmbeddingTitle: '云端嵌入模型服务：',
+          cloudEmbedding1: '检查API密钥是否正确配置',
+          cloudEmbedding2: '验证端点地址是否可访问',
+          cloudEmbedding3: '检查网络连接和代理设置',
+          cloudEmbedding4: '确认模型名称是否与服务商支持的一致'
         }
       },
       panel3: {
@@ -824,7 +831,7 @@ export default {
     },
     about: {
       appDescription: '支持多模态AI智能搜索的跨平台本地桌面应用（Windows/MacOS/Linux），通过语音、文本、图像多种输入方式，支持本地Ollama和云端OpenAI兼容大模型，为知识工作者提供更智能的文件检索体验。',
-      tagline: '本地优先 · 云端可选 · 隐私安全',
+      tagline: '本地优先 · 云端可选 · 隐私可控',
       features: {
         voice: {
           title: '语音搜索',
@@ -836,12 +843,12 @@ export default {
         },
         semantic: {
           title: '语义搜索',
-          description: 'BGE-M3向量嵌入 + Ollama/云端大模型理解'
+          description: 'BGE-M3/云端嵌入向量 + Ollama/云端大模型理解'
         }
       },
       techHighlight: {
         title: '本地优先，云端可选',
-        description: '支持本地Ollama和OpenAI兼容云端大模型，灵活切换。本地文件和索引数据始终存储在本地，只有搜索查询会发送到云端服务（使用云端模型时）。'
+        description: '支持本地Ollama/BGE-M3和云端OpenAI兼容API，灵活切换。本地文件和索引数据始终存储在本地，仅搜索查询会发送到云端（使用云端模型时）。性能与隐私的权衡由您选择。'
       }
     }
   }

frontend/src/renderer/src/views/Help.vue

@@ -182,18 +182,38 @@
 
             <a-collapse-panel key="2" :header="t('help.faq.panel2.header')">
               <div class="faq-content">
-                <p><strong>{{ t('help.faq.panel2.content.checkTitle') }}</strong></p>
+                <!-- 本地 Ollama 服务 -->
+                <p><strong>{{ t('help.faq.panel2.content.ollamaTitle') }}</strong></p>
                 <ul>
-                  <li>{{ t('help.faq.panel2.content.check1') }}</li>
-                  <li>{{ t('help.faq.panel2.content.check2') }}</li>
-                  <li>{{ t('help.faq.panel2.content.check3') }}</li>
-                  <li>{{ t('help.faq.panel2.content.check4') }}</li>
+                  <li>{{ t('help.faq.panel2.content.ollama1') }}</li>
+                  <li>{{ t('help.faq.panel2.content.ollama2') }}</li>
+                  <li>{{ t('help.faq.panel2.content.ollama3') }}</li>
+                  <li>{{ t('help.faq.panel2.content.ollama4') }}</li>
                 </ul>
-                <p><strong>{{ t('help.faq.panel2.content.commandsTitle') }}</strong></p>
+                <p><strong>{{ t('help.faq.panel2.content.ollamaCommandsTitle') }}</strong></p>
                 <ul>
-                  <li>{{ t('help.faq.panel2.content.command1') }}<code>ollama pull qwen2.5:1.5b</code></li>
-                  <li>{{ t('help.faq.panel2.content.command2') }}<code>ollama list</code></li>
-                  <li>{{ t('help.faq.panel2.content.command3') }}<code>ollama run qwen2.5:1.5b</code></li>
+                  <li>{{ t('help.faq.panel2.content.ollamaCommand1') }}<code>ollama pull qwen2.5:1.5b</code></li>
+                  <li>{{ t('help.faq.panel2.content.ollamaCommand2') }}<code>ollama list</code></li>
+                  <li>{{ t('help.faq.panel2.content.ollamaCommand3') }}<code>ollama run qwen2.5:1.5b</code></li>
+                </ul>
+
+                <!-- 云端大模型服务 -->
+                <p><strong>{{ t('help.faq.panel2.content.cloudLlmTitle') }}</strong></p>
+                <ul>
+                  <li>{{ t('help.faq.panel2.content.cloudLlm1') }}</li>
+                  <li>{{ t('help.faq.panel2.content.cloudLlm2') }}</li>
+                  <li>{{ t('help.faq.panel2.content.cloudLlm3') }}</li>
+                  <li>{{ t('help.faq.panel2.content.cloudLlm4') }}</li>
+                  <li>{{ t('help.faq.panel2.content.cloudLlm5') }}</li>
+                </ul>
+
+                <!-- 云端嵌入模型服务 -->
+                <p><strong>{{ t('help.faq.panel2.content.cloudEmbeddingTitle') }}</strong></p>
+                <ul>
+                  <li>{{ t('help.faq.panel2.content.cloudEmbedding1') }}</li>
+                  <li>{{ t('help.faq.panel2.content.cloudEmbedding2') }}</li>
+                  <li>{{ t('help.faq.panel2.content.cloudEmbedding3') }}</li>
+                  <li>{{ t('help.faq.panel2.content.cloudEmbedding4') }}</li>
                 </ul>
               </div>
             </a-collapse-panel>
@@ -319,6 +339,8 @@
                 <a-tag color="orange">CN-CLIP</a-tag>
                 <a-tag color="purple">BGE-M3</a-tag>
                 <a-tag color="cyan">OpenAI API</a-tag>
+                <a-tag color="magenta">Cloud Embedding</a-tag>
+                <a-tag color="volcano">MCP & Agent Skills</a-tag>
               </div>
             </a-card>
           </div>