1. 完善了服务端web部分，使用Vue编写(见serve_view)

2. 增加了对嵌入式设备完全自定义控制的功能

README.md

@@ -82,4 +82,8 @@
 
 本项目当前并不完善，还有很多需要优化的地方，并且尚未接入Yosuga_embedded。
 
-欢迎大家为本项目贡献代码。
+欢迎大家为本项目贡献代码。
+
+## Star History
+
+[![Star History Chart](https://api.star-history.com/svg?repos=Misakityan/Yosuga_server&type=Date&t=20260202)](https://www.star-history.com/#Misakityan/Yosuga_server&Date)

Test/embeddedServerTest.py

@@ -0,0 +1,269 @@
+
+import socket
+import struct
+import threading
+import json
+import os
+from typing import Optional
+
+from openai import OpenAI  # 确保 pip install openai
+
+from src.server_core.yosuga_embedded_server.device_manager import DeviceManager, DeviceInfo
+from src.server_core.yosuga_embedded_server.function_registry import FunctionRegistry
+from src.server_core.yosuga_embedded_server.ai_prompt import AIPromptBuilder
+from src.server_core.yosuga_embedded_server.json_rpc import JSONRPCHandler, RPCError
+
+
+class DeviceConnection:
+    """管理一个 TCP 设备连接"""
+    def __init__(self, sock, addr, server):
+        self.sock = sock
+        self.addr = addr
+        self.server = server
+        self.device_id: Optional[str] = None
+
+    def send_msg(self, data: str):
+        """发送带长度前缀的消息"""
+        encoded = data.encode('utf-8')
+        self.sock.sendall(struct.pack('!I', len(encoded)))
+        self.sock.sendall(encoded)
+
+    def recv_msg(self) -> Optional[str]:
+        try:
+            # 读满 4 字节长度前缀
+            raw_len = b''
+            while len(raw_len) < 4:
+                chunk = self.sock.recv(4 - len(raw_len))
+                if not chunk:  # 对方关闭连接
+                    return None
+                raw_len += chunk
+            msg_len = struct.unpack('!I', raw_len)[0]
+
+            # 读消息体
+            data = b''
+            while len(data) < msg_len:
+                chunk = self.sock.recv(msg_len - len(data))
+                if not chunk:
+                    return None
+                data += chunk
+            return data.decode('utf-8')
+        except Exception as e:
+            print(f"recv error: {e}")
+            return None
+
+    def handle(self):
+        try:
+            # 只接收第一条能力广告
+            caps_str = self.recv_msg()
+            if not caps_str:
+                return
+            print(f"[{self.addr}] Capabilities received")
+            device = self.server.register_device(caps_str)
+            self.device_id = device.device_id
+            self.server.set_connection(device.device_id, self)
+            print(f"[{self.addr}] Registered as {device.name} (id={device.device_id})")
+
+            # 此处直接返回，线程结束。后续通信由 call_device 接管。
+        except Exception as e:
+            print(f"Registration error: {e}")
+        finally:
+            # 注意：不要关闭 socket！它还要用于后续 call_device
+            pass
+
+
+class YosugaServer:
+    """整合设备管理、函数注册、AI 交互的服务端"""
+    def __init__(self, deepseek_api_key: str):
+        self.device_manager = DeviceManager()
+        self.function_registry = FunctionRegistry()
+        self.ai_prompt = AIPromptBuilder()
+        self._lock = threading.Lock()
+        self._call_id_counter = 0
+
+        # 关联设备变更
+        self.device_manager.on_device_change = self._on_device_change
+        self.function_registry.on_change = self._on_functions_change
+
+        # DeepSeek 客户端
+        self.ai_client = OpenAI(
+            api_key=deepseek_api_key,
+            base_url="https://api.deepseek.com"
+        )
+
+        # 设备连接表：device_id -> DeviceConnection
+        self._connections: dict[str, DeviceConnection] = {}
+
+    def _on_device_change(self, event: str, device: DeviceInfo):
+        print(f"Device event: {event} {device.device_id}")
+        if event == "removed":
+            self.function_registry.remove_device_functions(device.device_id)
+            conn = self._connections.pop(device.device_id, None)
+            if conn and conn.sock:
+                try:
+                    conn.sock.close()
+                except Exception:
+                    pass
+        elif event in ("added", "updated"):
+            if device.state.value == "registered":
+                self.function_registry.add_device_functions(
+                    device.device_id,
+                    device.name,
+                    device.functions or [],
+                )
+
+    def _on_functions_change(self):
+        print(f"Functions updated, total: {self.function_registry.function_count()}")
+
+    def register_device(self, caps_json: str) -> DeviceInfo:
+        data = json.loads(caps_json)
+        return self.device_manager.register_from_json(data)
+
+    def remove_device(self, device_id: str):
+        self.device_manager.remove_device(device_id)
+
+    def set_connection(self, device_id: str, conn: DeviceConnection):
+        with self._lock:
+            self._connections[device_id] = conn
+
+    def call_device(self, device_id: str, method: str, params: dict, call_id: int) -> dict:
+        conn = self._connections.get(device_id)
+        if not conn:
+            return {"error": {"code": RPCError.DEVICE_NOT_FOUND,
+                              "message": f"Device {device_id} not connected"}}
+
+        req = JSONRPCHandler.build_call(method, params, call_id)
+        try:
+            conn.send_msg(req)
+            resp_str = conn.recv_msg_with_timeout(timeout=5.0)
+            if resp_str is None:
+                # 超时或连接断开
+                self.remove_device(device_id)  # 自动清理
+                return {"error": {"code": RPCError.TIMEOUT, "message": "Device timeout or disconnected"}}
+
+            resp = JSONRPCHandler.parse_response(resp_str)
+            if resp and resp.is_success():
+                return {"result": resp.result}
+            elif resp and resp.error:
+                return {"error": resp.error.to_dict()}
+            else:
+                return {"error": {"code": RPCError.PARSE_ERROR, "message": "Invalid response"}}
+        except Exception as e:
+            self.remove_device(device_id)
+            return {"error": {"code": RPCError.DEVICE_ERROR, "message": str(e)}}
+
+    def process_device_message(self, device_id: str, message: str) -> str:
+        """处理从设备主动发来的消息（如 RPC 响应）"""
+        # 在这个架构中，设备不会主动发消息，所有交互都由服务端发起
+        # 这里仅作为占位，返回空
+        return ""
+
+    def handle_user_request(self, user_input: str) -> str:
+        """处理用户自然语言请求：调用 AI，执行函数，返回最终答案"""
+        # 1. 构建系统提示（包含当前所有函数）
+        func_list = self.function_registry.to_function_list()
+        system_prompt = self.ai_prompt.build_system_prompt(func_list)
+
+        # 2. 调用 DeepSeek
+        print("Calling DeepSeek...")
+        try:
+            response = self.ai_client.chat.completions.create(
+                model="deepseek-chat",
+                messages=[
+                    {"role": "system", "content": system_prompt},
+                    {"role": "user", "content": user_input}
+                ],
+                temperature=0.0,  # 确保输出稳定
+            )
+            ai_text = response.choices[0].message.content.strip()
+            print(f"AI response:\n{ai_text}")
+        except Exception as e:
+            return f"AI error: {e}"
+
+        # 3. 解析 AI 返回的调用
+        calls = self.ai_prompt.parse_ai_response(ai_text)
+        if not calls:
+            return "Failed to parse AI response as JSON-RPC calls"
+
+        # 4. 逐个执行调用
+        results = []
+        for idx, call in enumerate(calls):
+            method = call.get("method")
+            params = call.get("params", {})
+            call_id = idx + 1  # 简单自增 ID
+
+            # 找到该函数所属设备
+            func_info = self.function_registry.get_function(method)
+            if not func_info:
+                results.append(f"❌ Unknown function: {method}")
+                continue
+
+            device_id = func_info.device_id
+            print(f"Routing '{method}' to device {device_id}")
+
+            # 发送给设备
+            dev_resp = self.call_device(device_id, method, params, call_id)
+            if "error" in dev_resp:
+                results.append(f"❌ {method}: {dev_resp['error']['message']}")
+            else:
+                results.append(f"✅ {method}: {dev_resp.get('result', 'ok')}")
+
+        # 5. 汇总结果返回
+        summary = "\n".join(results)
+        print(f"Task result:\n{summary}")
+        return summary
+
+
+# ========== 修改 DeviceConnection 的 recv_msg 支持超时 ==========
+
+def recv_msg_with_timeout(self, timeout: float = 5.0) -> Optional[str]:
+    """接收带长度前缀的消息，带超时"""
+    self.sock.settimeout(timeout)
+    try:
+        return self.recv_msg()
+    except socket.timeout:
+        return None
+    finally:
+        self.sock.settimeout(None)
+
+DeviceConnection.recv_msg_with_timeout = recv_msg_with_timeout  # monkey-patch
+
+
+# ========== 主函数 ==========
+def main():
+
+
+    server = YosugaServer("")
+
+    # 启动 TCP 监听，接收设备连接
+    def accept_connections():
+        listener = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+        listener.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
+        listener.bind(('0.0.0.0', 9555))
+        listener.listen(5)
+        print("Server listening on port 9555...")
+        while True:
+            sock, addr = listener.accept()
+            print(f"New connection from {addr}")
+            conn = DeviceConnection(sock, addr, server)
+            # 在连接注册后，关联 device_id 需要等到能力广告，暂存一下
+            # 我们在 handle 中收到能力后再设置 server.set_connection
+            threading.Thread(target=conn.handle, daemon=True).start()
+
+    threading.Thread(target=accept_connections, daemon=True).start()
+
+    # 简单交互循环
+    print("\nEnter your requests (type 'quit' to exit):")
+    while True:
+        try:
+            user_input = input("> ")
+        except EOFError:
+            break
+        if user_input.lower() in ('quit', 'exit'):
+            break
+        if not user_input.strip():
+            continue
+        result = server.handle_user_request(user_input)
+        print("---\n" + result + "\n---")
+
+if __name__ == "__main__":
+    main()

main.py

@@ -41,5 +41,5 @@ if __name__ == "__main__":
     try:
         asyncio.run(main())
     except KeyboardInterrupt:
-        print("\nYosuga服务器已停止喵~~~")
+        print("\nYosuga服务端已停止喵~~~")
     

pyproject.toml

@@ -7,10 +7,15 @@ requires-python = ">=3.11"
 dependencies = [
     "aiofiles>=25.1.0",
     "aiohttp>=3.13.3",
+    "eventlet>=0.40.4",
     "fastapi>=0.128.0",
     "faster-whisper>=1.2.1",
+    "flask>=3.1.3",
+    "flask-cors>=6.0.2",
+    "flask-socketio>=5.6.1",
     "loguru>=0.7.3",
     "openai>=2.16.0",
+    "psutil>=7.2.2",
     "pyautogui>=0.9.54",
     "pydantic>=2.12.5",
     "pydantic-settings>=2.12.0",

src/config/config.py

@@ -29,7 +29,7 @@ class AIConfig:
     model_name: str = "qwen/qwen3-4b-2507"
     timeout: int = 30
     temperature: float = 0.4
-    max_tokens: int = 8192
+    max_tokens: int = 4096
 
 
 @dataclass
@@ -98,12 +98,22 @@ class AppConfig:
         # 基础字段
         self.version = version
         self.debug = debug
-        self.ai = ai if ai is not None else AIConfig()
-        self.tts = tts if tts is not None else TTSConfig()
-        self.asr = asr if asr is not None else ASRConfig()
-        self.auto_agent = auto_agent if auto_agent is not None else AutoAgentConfig()
-        self.llm_core = llm_core if llm_core is not None else LLMConfig()
-        self.paths = paths if paths is not None else PathsConfig()
+
+        # self.ai = ai if ai is not None else AIConfig()
+        # self.tts = tts if tts is not None else TTSConfig()
+        # self.asr = asr if asr is not None else ASRConfig()
+        # self.auto_agent = auto_agent if auto_agent is not None else AutoAgentConfig()
+        # self.llm_core = llm_core if llm_core is not None else LLMConfig()
+        # self.paths = paths if paths is not None else PathsConfig()
+
+        # 如果是字典则转换，否则使用默认值
+        self.ai = AIConfig(**ai) if isinstance(ai, dict) else (ai or AIConfig())
+        self.tts = TTSConfig(**tts) if isinstance(tts, dict) else (tts or TTSConfig())
+        self.asr = ASRConfig(**asr) if isinstance(asr, dict) else (asr or ASRConfig())
+        self.auto_agent = AutoAgentConfig(**auto_agent) if isinstance(auto_agent, dict) else (
+                    auto_agent or AutoAgentConfig())
+        self.llm_core = LLMConfig(**llm_core) if isinstance(llm_core, dict) else (llm_core or LLMConfig())
+        self.paths = PathsConfig(**paths) if isinstance(paths, dict) else (paths or PathsConfig())
 
         # 内部状态（非 dataclass 字段，不会被序列化）
         self._config_path = _config_path
@@ -395,6 +405,12 @@ class _LazyConfig:
     def using_dir(self) -> Path:
         return _ensure_initialized().using_dir
 
+def ensure_config_initialized():
+    """
+    强制立即初始化配置（用于多线程环境）
+    返回真正的 AppConfig 实例而非代理
+    """
+    return _ensure_initialized()
 
 # 全局配置对象：导入即用，自动初始化
 cfg: AppConfig = _LazyConfig()  # type: ignore

src/modules/asr_module/asr_core/fast_whisper/asr_interface.py

@@ -39,7 +39,7 @@ class ASRInterface:
         self.sample_rate = 16000
         
         self._initialized = True
-        logger.info("🎤 ASR接口初始化完成")
+        logger.info("ASR接口初始化完成")
     
     @classmethod
     def get_instance(cls, config: Optional[ASRConfig] = None) -> 'ASRInterface':
@@ -68,7 +68,7 @@ class ASRInterface:
             import time
             start_time = time.time()
             
-            logger.info(f"🎵 开始识别: {wav_path.name}")
+            logger.info(f"开始识别: {wav_path.name}")
             
             # 执行识别...
             audio = self._load_audio(wav_path)
@@ -78,14 +78,14 @@ class ASRInterface:
             # 计算耗时
             processing_time = time.time() - start_time
             logger.info(
-                f"✅ 识别完成: {lang} | {len(text)}字符 | 置信度:{confidence:.2f} | "
+                f"识别完成: {lang} | {len(text)}字符 | 置信度:{confidence:.2f} | "
                 f"耗时:{processing_time:.3f}s | RTF:{processing_time/(len(audio)/self.sample_rate):.3f}"
             )
             
             return text, lang, confidence
             
         except Exception as e:
-            logger.error(f"❌ 识别失败 {wav_path}: {e}")
+            logger.error(f"识别失败 {wav_path}: {e}")
             raise RuntimeError(f"Transcription failed: {e}")
     
     def _load_audio(self, wav_path: Path) -> numpy.ndarray:
@@ -180,5 +180,5 @@ class ASRInterface:
     
     def shutdown(self):
         """优雅关闭"""
-        logger.info("🛑 关闭ASR接口...")
+        logger.info("关闭ASR接口...")
         self.model_manager.unload()

src/modules/asr_module/asr_core/fast_whisper/model_manager.py

@@ -34,8 +34,8 @@ class ModelManager:
     
     def _load_model(self):
         """加载模型"""
-        logger.info(f"🚀 初始化模型: {self.config.model_name}")
-        logger.info(f"📦 设备: {self.config.device}, 计算类型: {self.config.compute_type}")
+        logger.info(f"初始化模型: {self.config.model_name}")
+        logger.info(f"设备: {self.config.device}, 计算类型: {self.config.compute_type}")
         
         try:
             self._model = WhisperModel(
@@ -52,15 +52,15 @@ class ModelManager:
                 "model_size": self.config.model_name.split("-")[-2]
             }
             
-            logger.info("✅ 模型加载成功")
+            logger.info("模型加载成功")
             
         except Exception as e:
-            logger.error(f"❌ 模型加载失败: {e}")
+            logger.error(f"模型加载失败: {e}")
             raise RuntimeError(f"Failed to load ASR model: {e}")
     
     def reload(self, new_config: ASRConfig):
         """热重载模型"""
-        logger.info("🔄 热重载模型...")
+        logger.info("热重载模型...")
         self.unload()
         self.config = new_config
         self._load_model()
@@ -68,7 +68,7 @@ class ModelManager:
     def unload(self):
         """卸载模型释放资源"""
         if self._model is not None:
-            logger.info("🗑️ 卸载模型...")
+            logger.info("卸载模型...")
             del self._model
             self._model = None
             
@@ -77,7 +77,7 @@ class ModelManager:
             
             gc.collect()
             
-            logger.info("✅ 模型已卸载")
+            logger.info("模型已卸载")
     
     def get_device_info(self) -> dict:
         """获取设备信息"""

src/modules/asr_module/asr_core/fast_whisper/utils.py

@@ -42,4 +42,4 @@ class PerformanceProfiler:
         
         if len(self.stats) % 10 == 0:
             avg_rtf = sum(s["rtf"] for s in self.stats[-10:]) / 10
-            logger.info(f"📊 最近10次平均RTF: {avg_rtf:.3f}")
+            logger.info(f"最近10次平均RTF: {avg_rtf:.3f}")

src/modules/asr_module/client/asr_client.py

@@ -82,7 +82,7 @@ class ASRClientSync:
         if not file_path.exists():
             raise FileNotFoundError(f"文件不存在: {file_path}")
 
-        logger.info(f"📤 上传文件: {file_path.name}")
+        logger.info(f"上传文件: {file_path.name}")
 
         with open(file_path, 'rb') as f:
             files = {'file': (file_path.name, f, 'audio/wav')}
@@ -106,7 +106,7 @@ class ASRClientSync:
                 audio_bytes = f.read()
             result = client.transcribe_bytes(audio_bytes)
         """
-        logger.info(f"📤 上传字节流 ({len(audio_data)} bytes)")
+        logger.info(f"上传字节流 ({len(audio_data)} bytes)")
 
         files = {'file': (filename, audio_data, 'audio/wav')}
         result = self._request('POST', '/transcribe', files=files)
@@ -171,7 +171,7 @@ class ASRClientAsync:
         if not file_path.exists():
             raise FileNotFoundError(f"文件不存在: {file_path}")
 
-        logger.info(f"📤 上传文件: {file_path.name}")
+        logger.info(f"上传文件: {file_path.name}")
 
         async with aiofiles.open(file_path, 'rb') as f:
             audio_data = await f.read()
@@ -180,7 +180,7 @@ class ASRClientAsync:
 
     async def transcribe_bytes(self, audio_data: bytes, filename: str = "audio.wav") -> ASRResponse:
         """异步转录音频字节流"""
-        logger.info(f"📤 上传字节流 ({len(audio_data)} bytes)")
+        logger.info(f"上传字节流 ({len(audio_data)} bytes)")
         await self._ensure_session()    # 确保session已创建
         form = aiohttp.FormData()       # 创建表单数据
         form.add_field('file', audio_data, filename=filename, content_type='audio/wav') # 添加文件字段

src/modules/asr_module/start_api.py

@@ -20,7 +20,7 @@ def first_test() -> None:
     """首次启动测试"""
     time.sleep(5)   # 给服务器一些启动时间
     # 构造一个测试请求以验证初始化模型加载成功
-    logger.info("🚀 测试模型是否加载成功...")
+    logger.info("测试模型是否加载成功...")
     import requests
     from pathlib import Path
     url = "http://localhost:20260/transcribe"
@@ -52,7 +52,7 @@ def first_test() -> None:
         logger.error(f"测试过程中发生错误: {e}")
 
 if __name__ == "__main__":
-    logger.info("🚀 启动 ASR API 服务...")
+    logger.info("启动 ASR API 服务...")
 
     # 在后台线程启动服务器
     server_thread = threading.Thread(target=start_server, daemon=True)

src/modules/text_ai_module/text_ai_core/general_text_ai_req.py

@@ -234,7 +234,7 @@ class OpenAIClient(BaseLLMClient):
 
     def _normal_chat_completion(self, params):
         """非流式响应处理"""
-        logger.info("📡 发送非流式请求...")
+        logger.info("发送非流式请求...")
         response = self.client.chat.completions.create(**params)
         raw_usage = response.usage
         normalized_usage = normalize_usage(
@@ -251,7 +251,7 @@ class OpenAIClient(BaseLLMClient):
 
     def _stream_chat_completion(self, params):
         """流式响应处理"""
-        logger.info("📡 发送流式请求...")
+        logger.info("发送流式请求...")
         response_stream = self.client.chat.completions.create(**params)
 
         full_content = ""

src/modules/tts_module/tts_core/async_audio_player.py

@@ -49,7 +49,7 @@ class AsyncAudioPlayer:
                         # 读取PCM数据（去掉头部）
                         pcm_data = wav_file.readframes(wav_file.getnframes())
 
-                        logger.info(f"📊 解析WAV头: {self.sample_rate}Hz, {self.channels}ch, {self.sampwidth * 8}bit")
+                        logger.info(f"解析WAV头: {self.sample_rate}Hz, {self.channels}ch, {self.sampwidth * 8}bit")
 
                         # 转换为numpy数组
                         if self.sampwidth == 2:
@@ -68,7 +68,7 @@ class AsyncAudioPlayer:
 
                 except wave.Error:
                     # 可能是不完整的WAV头，尝试直接播放
-                    logger.warning("⚠️ WAV头解析失败，尝试直接播放")
+                    logger.warning("WAV头解析失败，尝试直接播放")
                     await self._play_raw(audio_data)
                     return
             else:
@@ -76,7 +76,7 @@ class AsyncAudioPlayer:
                 await self._play_raw(audio_data)
 
         except Exception as e:
-            logger.error(f"❌ 音频块处理失败: {e}")
+            logger.error(f"音频块处理失败: {e}")
 
     async def _play_raw(self, audio_data: bytes):
         """播放RAW PCM数据"""
@@ -90,11 +90,11 @@ class AsyncAudioPlayer:
 
             await self.audio_queue.put(audio_array)
         except Exception as e:
-            logger.error(f"❌ RAW音频处理失败: {e}")
+            logger.error(f"RAW音频处理失败: {e}")
 
     async def play_worker(self):
         """后台播放任务"""
-        logger.info("🎧 音频播放任务启动")
+        logger.info("音频播放任务启动")
 
         while self.is_playing or not self.audio_queue.empty():
             try:
@@ -103,7 +103,7 @@ class AsyncAudioPlayer:
 
                 # 延迟初始化音频流（直到获得第一个数据块）
                 if self.stream is None:
-                    logger.info(f"🔊 打开音频输出流: {self.sample_rate}Hz")
+                    logger.info(f"打开音频输出流: {self.sample_rate}Hz")
                     self.stream = sd.OutputStream(
                         samplerate=self.sample_rate,
                         channels=1,
@@ -122,10 +122,10 @@ class AsyncAudioPlayer:
             except asyncio.TimeoutError:
                 continue
             except Exception as e:
-                logger.error(f"❌ 播放任务异常: {e}")
+                logger.error(f"播放任务异常: {e}")
                 break
 
-        logger.info("🛑 音频播放任务结束")
+        logger.info("音频播放任务结束")
 
     async def start(self):
         """启动播放系统"""
@@ -154,7 +154,7 @@ class AsyncAudioPlayer:
             except:
                 break
 
-        logger.info("✅ 音频播放已停止")
+        logger.info("音频播放已停止")
 
     async def __aenter__(self):
         await self.start()

src/modules/websocket_base_module/dto/dto_templates/device_command_dto.py

@@ -0,0 +1,19 @@
+"""
+设备命令 DTO - 服务端向客户端发送嵌入式设备控制指令
+"""
+
+from pydantic import Field, BaseModel
+from datetime import datetime, timezone
+
+
+class DeviceCommandDataTransferObject(BaseModel):
+    """设备命令数据传输对象"""
+    device_id: str = Field(default="", description="目标设备ID")
+    payload: str = Field(default="", description="JSON-RPC 调用字符串")
+
+    def to_json(self) -> dict:
+        return {
+            "type": "device_command",
+            "timestamp": datetime.now(timezone.utc).timestamp(),
+            "data": self.model_dump()
+        }

src/modules/websocket_base_module/dto/second_dtos.py

@@ -38,7 +38,9 @@ class JsonDTO(MessageDTO):
         super().__init__(ws_server)
         self.receivers : Dict[str, List[ReceiveCallback]] = {
             'audio_data' : [],      # 音频数据
-            'screenshot_data' : []  # 截图数据
+            'screenshot_data' : [], # 截图数据
+            'device_data' : [],     # 嵌入式设备数据（注册/响应/事件）
+            'device_command' : [],  # 设备控制指令（服务端→客户端）
         }
         # 注册json处理callback function
         ws_server.register_receiver('json', self._handle_json)

src/server_core/core.py

@@ -25,6 +25,7 @@
 import asyncio
 from typing import Optional, List, Dict, Any
 from loguru import logger
+import json
 from src.modules.websocket_base_module.dto.third_dtos import (
     AudioDataDTO, AudioDataTransferObject,
     ScreenShotDataDTO, ScreenShotDataTransferObject
@@ -43,9 +44,15 @@ from src.server_core.llm_core.llm_core import (
     YosugaLLMCore, ModelProvider,
     LLMCoreAnalysisBase,
     YosugaAudioResponseData, YosugaUITARSResponseData,
-    YosugaUITARSRequestData
+    YosugaUITARSRequestData, YosugaEmbeddedResponseData
 )
 
+from src.server_core.yosuga_embedded_server import (
+    YosugaServer, ServerConfig
+)
+from src.server_core.yosuga_embedded_server.device_dto import DeviceDataDTO
+from src.server_core.llm_core.llm_core_prompt_manager import YosugaEmbedded
+
 from src.modules.websocket_base_module.dto.dto_templates.auto_agent_data_dto import AutoAgentDataTransferObject
 from src.config.config import cfg
 
@@ -70,6 +77,66 @@ class YosugaServerCore:
 
     llm_core: YosugaLLMCore = None      # llm core
 
+    embedded_server: YosugaServer       # 嵌入式设备管理框架
+    device_dto: DeviceDataDTO           # 设备数据分发器
+
+    # @classmethod
+    # async def get_instance(cls) -> "YosugaServerCore":
+    #     """异步单例工厂"""
+    #     if cls._instance is None:
+    #         async with cls._lock:
+    #             if cls._instance is None:
+    #                 logger.info("Initializing YosugaServerCore...")
+    #                 # 创建实例
+    #                 instance = cls.__new__(cls)
+    #
+    #                 # 按依赖顺序初始化数据分发器
+    #                 instance.ws_server = await get_ws_server()
+    #                 instance.json_dto = await get_json_dto_instance(instance.ws_server)
+    #                 instance.audio_dto = AudioDataDTO(instance.json_dto)            # 音频分发器
+    #                 instance.audio_dto.register_audio_callback(instance._handle_audio_data) # 注册音频处理函数
+    #                 instance.screenshot_dto = ScreenShotDataDTO(instance.json_dto)  # 截图分发器
+    #                 instance.screenshot_dto.register_screenshot_callback(instance._handle_screenshot_data)  # 注册截图处理函数
+    #
+    #                 instance.asr_client = create_asr_client(use_async=True, base_url=cfg.asr.url)
+    #                 instance.tts_client = GPTSoVITSClient(host=cfg.tts.host, port=cfg.tts.port, debug=True)
+    #                 # 切换GPT_SoVITS模型
+    #                 await instance.tts_client.set_gpt_weights(cfg.tts.gpt_model_name)
+    #                 await instance.tts_client.set_sovits_weights(cfg.tts.sovits_model_name)
+    #
+    #                 instance.auto_agent_client = UITarsClient(UITarsClientConfig(
+    #                     deployment_type=cfg.auto_agent.deployment_type,
+    #                     base_url=cfg.auto_agent.base_url,
+    #                     model_name=cfg.auto_agent.model_name,
+    #                     temperature=cfg.auto_agent.temperature,
+    #                     max_tokens=cfg.auto_agent.max_tokens
+    #                 ))
+    #
+    #                 instance.llm_core = YosugaLLMCore(
+    #                     model_config=ModelConfig(       # TODO 同上
+    #                         provider=ModelProvider.OPENAI,
+    #                         model_name=cfg.ai.model_name,
+    #                         base_url=cfg.ai.base_url,
+    #                         api_key=cfg.ai.api_key,
+    #                         temperature=cfg.ai.temperature,
+    #                         max_tokens=cfg.ai.max_tokens
+    #                     ),
+    #                     core_config=LLMCoreConfig(      # TODO 同上
+    #                         max_context_tokens=cfg.llm_core.max_context_tokens,
+    #                         enable_history=cfg.llm_core.enable_history,
+    #                         role_setting=cfg.llm_core.role_character,
+    #                         language=cfg.llm_core.language,      # 回复使用语言
+    #                         auto_dispatch=True,
+    #                         dispatch_async=True  # 启用异步分发
+    #                     )
+    #                 )
+    #                 instance.register_llm_core_analysis()   # 注册解析器
+    #                 instance.register_llm_core_action()     # 注册分发器
+    #                 instance.llm_core.register_overflow_handler(instance._handle_overflow_logger)   # 注册上下文溢出处理回调
+    #
+    #                 cls._instance = instance
+    #                 logger.success("YosugaServerCore initialized")
+    #     return cls._instance
     @classmethod
     async def get_instance(cls) -> "YosugaServerCore":
         """异步单例工厂"""
@@ -77,57 +144,141 @@ class YosugaServerCore:
             async with cls._lock:
                 if cls._instance is None:
                     logger.info("Initializing YosugaServerCore...")
+
+                    # 强制初始化配置
+                    from src.config.config import _ensure_initialized
+                    from dataclasses import asdict, is_dataclass
+
+                    real_cfg = _ensure_initialized()
+
+                    # 辅助函数：递归转换为 dict
+                    def to_dict(obj):
+                        if isinstance(obj, dict):
+                            return obj
+                        if is_dataclass(obj) and not isinstance(obj, type):
+                            return asdict(obj)
+                        return {}
+
+                    # 提取各个配置段并转换为 dict（关键修复）
+                    cfg_dict = {
+                        'ai': to_dict(getattr(real_cfg, 'ai', {})),
+                        'tts': to_dict(getattr(real_cfg, 'tts', {})),
+                        'asr': to_dict(getattr(real_cfg, 'asr', {})),
+                        'auto_agent': to_dict(getattr(real_cfg, 'auto_agent', {})),
+                        'llm_core': to_dict(getattr(real_cfg, 'llm_core', {})),
+                    }
+
+                    logger.debug(f"配置提取完成: ai={type(cfg_dict['ai'])}, tts={type(cfg_dict['tts'])}")
+
                     # 创建实例
                     instance = cls.__new__(cls)
 
                     # 按依赖顺序初始化数据分发器
                     instance.ws_server = await get_ws_server()
                     instance.json_dto = await get_json_dto_instance(instance.ws_server)
-                    instance.audio_dto = AudioDataDTO(instance.json_dto)            # 音频分发器
-                    instance.audio_dto.register_audio_callback(instance._handle_audio_data) # 注册音频处理函数
-                    instance.screenshot_dto = ScreenShotDataDTO(instance.json_dto)  # 截图分发器
-                    instance.screenshot_dto.register_screenshot_callback(instance._handle_screenshot_data)  # 注册截图处理函数
+                    instance.audio_dto = AudioDataDTO(instance.json_dto)
+                    instance.audio_dto.register_audio_callback(instance._handle_audio_data)
+                    instance.screenshot_dto = ScreenShotDataDTO(instance.json_dto)
+                    instance.screenshot_dto.register_screenshot_callback(instance._handle_screenshot_data)
 
-                    instance.asr_client = create_asr_client(use_async=True, base_url=cfg.asr.url)
-                    instance.tts_client = GPTSoVITSClient(host=cfg.tts.host, port=cfg.tts.port, debug=True)
-                    # 切换GPT_SoVITS模型
-                    await instance.tts_client.set_gpt_weights(cfg.tts.gpt_model_name)
-                    await instance.tts_client.set_sovits_weights(cfg.tts.sovits_model_name)
+                    # ASR 客户端
+                    asr_cfg = cfg_dict.get('asr', {})
+                    instance.asr_client = create_asr_client(
+                        use_async=True,
+                        base_url=asr_cfg.get('url', 'http://localhost:20260/')
+                    )
 
+                    # TTS 客户端
+                    tts_cfg = cfg_dict.get('tts', {})
+                    instance.tts_client = GPTSoVITSClient(
+                        host=tts_cfg.get('host', 'localhost'),
+                        port=tts_cfg.get('port', 20261),
+                        debug=True
+                    )
+
+                    # 切换 GPT_SoVITS 模型
+                    # await instance.tts_client.set_gpt_weights(
+                    #     tts_cfg.get('gpt_model_name', 'GPT_weights_v2Pro/Yosuga_Airi-e32.ckpt')
+                    # )
+                    # await instance.tts_client.set_sovits_weights(
+                    #     tts_cfg.get('sovits_model_name', 'SoVITS_weights_v2Pro/Yosuga_Airi_e16_s864.pth')
+                    # )
+
+                    # Auto Agent 客户端
+                    auto_cfg = cfg_dict.get('auto_agent', {})
                     instance.auto_agent_client = UITarsClient(UITarsClientConfig(
-                        deployment_type=cfg.auto_agent.deployment_type,
-                        base_url=cfg.auto_agent.base_url,
-                        model_name=cfg.auto_agent.model_name,
-                        temperature=cfg.auto_agent.temperature,
-                        max_tokens=cfg.auto_agent.max_tokens
+                        deployment_type=auto_cfg.get('deployment_type', 'lmstudio'),
+                        base_url=auto_cfg.get('base_url', 'http://localhost:1234/v1'),
+                        model_name=auto_cfg.get('model_name', 'ui-tars-1.5-7b@q4_k_m'),
+                        temperature=auto_cfg.get('temperature', 0.1),
+                        max_tokens=auto_cfg.get('max_tokens', 16384)
                     ))
 
+                    # LLM Core
+                    ai_cfg = cfg_dict.get('ai', {})
+                    llm_cfg = cfg_dict.get('llm_core', {})
+
                     instance.llm_core = YosugaLLMCore(
-                        model_config=ModelConfig(       # TODO 同上
+                        model_config=ModelConfig(
                             provider=ModelProvider.OPENAI,
-                            model_name=cfg.ai.model_name,
-                            base_url=cfg.ai.base_url,
-                            api_key=cfg.ai.api_key,
-                            temperature=cfg.ai.temperature,
-                            max_tokens=cfg.ai.max_tokens
+                            model_name=ai_cfg.get('model_name', 'qwen/qwen3-4b-2507'),
+                            base_url=ai_cfg.get('base_url', 'http://localhost:1234/v1'),
+                            api_key=ai_cfg.get('api_key'),
+                            temperature=ai_cfg.get('temperature', 0.4),
+                            max_tokens=ai_cfg.get('max_tokens', 8192)
                         ),
-                        core_config=LLMCoreConfig(      # TODO 同上
-                            max_context_tokens=cfg.llm_core.max_context_tokens,
-                            enable_history=cfg.llm_core.enable_history,
-                            role_setting=cfg.llm_core.role_character,
-                            language=cfg.llm_core.language,      # 回复使用语言
+                        core_config=LLMCoreConfig(
+                            max_context_tokens=llm_cfg.get('max_context_tokens', 2048),
+                            enable_history=llm_cfg.get('enable_history', True),
+                            role_setting=llm_cfg.get('role_character',
+                                                     '你是由Misakiotoha开发的助手稲葉愛理ちゃん，可以和用户一起玩游戏，聊天，做各种事情，性格抽象，没事爱整整活。'),
+                            language=llm_cfg.get('language', '中文'),
                             auto_dispatch=True,
-                            dispatch_async=True  # 启用异步分发
+                            dispatch_async=True
                         )
                     )
-                    instance.register_llm_core_analysis()   # 注册解析器
-                    instance.register_llm_core_action()     # 注册分发器
-                    instance.llm_core.register_overflow_handler(instance._handle_overflow_logger)   # 注册上下文溢出处理回调
+
+                    # 注册 YosugaEmbedded 提示词模块
+                    instance.llm_core.register_prompt_module(YosugaEmbedded())
+                    logger.info("[Core] 嵌入式设备提示词模块已注册")
+
+                    # 初始化嵌入式设备管理框架
+                    instance.embedded_server = YosugaServer(
+                        config=ServerConfig(
+                            device_conflict_strategy="rename",
+                            max_concurrent_calls=10,
+                            device_timeout=30.0,
+                        )
+                    )
+                    instance.device_dto = DeviceDataDTO(
+                        instance.json_dto, instance.embedded_server
+                    )
+                    # 当 YosugaServer 需要发送 RPC 到设备时，通过 WebSocket 发出 device_command
+                    instance.embedded_server.on_device_message = (
+                        instance._on_device_message
+                    )
+                    # 当设备能力变更时，更新 LLM 系统提示词中的状态表
+                    instance.embedded_server.on_capabilities_changed = (
+                        instance._on_capabilities_changed
+                    )
+                    logger.success("[Core] 嵌入式设备管理框架已初始化")
+
+                    # 注册设备 RPC 响应回调（设备结果回来后喂回 LLM）
+                    instance.device_dto.register_device_callback(
+                        instance._on_device_rpc_response
+                    )
+                    instance._pending_rpc: Optional[dict] = None
+
+                    instance.register_llm_core_analysis()
+                    instance.register_llm_core_action()
+                    instance.llm_core.register_overflow_handler(instance._handle_overflow_logger)
 
                     cls._instance = instance
                     logger.success("YosugaServerCore initialized")
+
         return cls._instance
 
+
     def register_llm_core_action(self):
         """
         注册llm_core的分发器
@@ -137,6 +288,7 @@ class YosugaServerCore:
         self.llm_core.register_action_handler("audio_text", self._handle_audio_response, is_async=True)
         self.llm_core.register_action_handler("auto_agent", self._handle_auto_agent, is_async=True)
         self.llm_core.register_action_handler("call_auto_agent", self._handle_call_auto_agent, is_async=True)
+        self.llm_core.register_action_handler("embedded_control", self._handle_embedded_control, is_async=True)
         self.llm_core.set_fallback_handler(self._handle_fallback)
 
     def register_llm_core_analysis(self):
@@ -148,6 +300,7 @@ class YosugaServerCore:
         self.llm_core.register_analysis_model(YosugaAudioResponseData)
         self.llm_core.register_analysis_model(YosugaUITARSResponseData)
         self.llm_core.register_analysis_model(YosugaUITARSRequestData)
+        self.llm_core.register_analysis_model(YosugaEmbeddedResponseData)
 
     def _handle_overflow_logger(self, history: List[Any], metadata: Dict[str, Any]):
         """上下文溢出记录，仅打印日志"""
@@ -211,14 +364,23 @@ class YosugaServerCore:
             try:
                 # 使用最快模式流式输出
                 chunk_count = 0
+                # async for chunk in await self.tts_client.tts(
+                #         text=data.response_text,
+                #         ref_audio_path="uploaded_audio/test_voice.wav", # TODO 需要替换成config或者后续设计情感系统
+                #         text_lang="ja",
+                #         prompt_lang="ja",
+                #         prompt_text="もう!こんなところで何やってるんだよ!",  # 参考语音的真实文本
+                #         streaming_mode=StreamingMode.FASTEST,  # 模式3：快速流式
+                #         media_type="wav"
+                # ):
                 async for chunk in await self.tts_client.tts(
-                        text=data.response_text,
-                        ref_audio_path="uploaded_audio/test_voice.wav", # TODO 需要替换成config或者后续设计情感系统
-                        text_lang="ja",
-                        prompt_lang="ja",
-                        prompt_text="もう!こんなところで何やってるんだよ!",  # 参考语音的真实文本
-                        streaming_mode=StreamingMode.FASTEST,  # 模式3：快速流式
-                        media_type="wav"
+                    text=data.response_text,
+                    ref_audio_path="uploaded_audio/kq.wav", # TODO 需要替换成config或者后续设计情感系统
+                    text_lang="zh",
+                    prompt_lang="zh",
+                    prompt_text="电闪雷鸣虽然有点吓人，但璃月港的防雷防火工事是一流的，不用担心。",  # 参考语音的真实文本
+                    streaming_mode=StreamingMode.FASTEST,  # 模式3：快速流式
+                    media_type="wav"
                 ):
                     chunk_count += 1
                     # print(f"🎵 收到音频块 #{chunk_count}: {len(chunk.audio_data)} bytes")
@@ -245,7 +407,7 @@ class YosugaServerCore:
                                 text=data.response_text
                             )
                         )
-                print(f"✅ 流式TTS完成！共{chunk_count}个音频块")
+                print(f"流式TTS完成！共{chunk_count}个音频块")
                 # 构造音频尾包发送给客户端(虚假的音频数据)
                 await self.audio_dto.send_audio_data(
                     AudioDataTransferObject(
@@ -258,7 +420,7 @@ class YosugaServerCore:
                     )
                 )
             except Exception as e:
-                print(f"❌ 流式错误: {e}")
+                print(f"流式错误: {e}")
             return {"status": "success", "executed": data.response_text}
         return None
 
@@ -284,6 +446,125 @@ class YosugaServerCore:
             await self.screenshot_dto.send_screenshot_data(ScreenShotDataTransferObject(LLMResponse=data.llm_translation))
         return {"status": "success", "executed": data.type}
 
+    async def _handle_embedded_control(self, data: YosugaEmbeddedResponseData):
+        """
+        llm_core异步处理器：嵌入式设备控制
+        将LLM输出的 JSON-RPC 调用列表交由 YosugaServer 框架处理并路由到对应设备
+        """
+        logger.info(f"Handling embedded control: {len(data.calls)} calls")
+
+        results = self.embedded_server.process_ai_response(json.dumps(data.calls))
+        logger.info(f"Embedded control results: {results}")
+
+        # 保存 pending RPC 信息，等设备异步响应回来后喂回 LLM
+        if results and len(results) > 0:
+            first_call = results[0]
+            self._pending_rpc = {
+                "device_id": first_call.get("device_id"),
+                "method": first_call.get("method"),
+                "call_id": first_call.get("id"),
+                "original_response_text": data.response_text or "",
+            }
+
+        # 如果 LLM 同时返回了需要回复用户的文本，通过 TTS 播报
+        if data.response_text:
+            try:
+                chunk_count = 0
+                # async for chunk in await self.tts_client.tts(
+                #     text=data.response_text,
+                #     ref_audio_path="uploaded_audio/test_voice.wav",
+                #     text_lang="ja",
+                #     prompt_lang="ja",
+                #     prompt_text="もう!こんなところで何やってるんだよ!",
+                #     streaming_mode=StreamingMode.FASTEST,
+                #     media_type="wav"
+                # ):
+                async for chunk in await self.tts_client.tts(
+                    text=data.response_text,
+                    ref_audio_path="uploaded_audio/kq.wav", # TODO 需要替换成config或者后续设计情感系统
+                    text_lang="zh",
+                    prompt_lang="zh",
+                    prompt_text="电闪雷鸣虽然有点吓人，但璃月港的防雷防火工事是一流的，不用担心。",  # 参考语音的真实文本
+                    streaming_mode=StreamingMode.FASTEST,  # 模式3：快速流式
+                    media_type="wav"
+                ):
+                    chunk_count += 1
+                    if chunk_count == 1:
+                        await self.audio_dto.send_audio_data(
+                            AudioDataTransferObject(
+                                data=chunk.audio_data,
+                                isStream=True, isStart=True,
+                                sequence=chunk_count, isEnd=False,
+                                text=data.response_text
+                            )
+                        )
+                    else:
+                        await self.audio_dto.send_audio_data(
+                            AudioDataTransferObject(
+                                data=chunk.audio_data,
+                                isStream=True, isStart=False,
+                                sequence=chunk_count, isEnd=False,
+                                text=data.response_text
+                            )
+                        )
+                await self.audio_dto.send_audio_data(
+                    AudioDataTransferObject(
+                        data=b"0",
+                        isStream=True, isStart=False,
+                        sequence=chunk_count + 1, isEnd=True,
+                        text=data.response_text
+                    )
+                )
+            except Exception as e:
+                logger.error(f"Embedded control TTS error: {e}")
+
+        return {"status": "success", "calls": len(data.calls)}
+
+    def _on_device_rpc_response(self, device_id: str, payload: dict):
+        """DeviceDataDTO 回调：设备 RPC 响应回来时触发，喂回 LLM"""
+        if self._pending_rpc and self._pending_rpc.get("device_id") == device_id:
+            call_id = payload.get("id")
+            if call_id is None or call_id == self._pending_rpc.get("call_id"):
+                pending = self._pending_rpc
+                self._pending_rpc = None
+                asyncio.create_task(self._continue_with_device_result(device_id, payload, pending))
+
+    async def _continue_with_device_result(self, device_id: str, payload: dict, pending: dict):
+        """设备 RPC 结果回来后，喂回 LLM 生成最终回复并 TTS"""
+        method = pending.get("method", "unknown")
+        original_text = pending.get("original_response_text", "")
+
+        result_str = json.dumps(payload.get("result", payload), ensure_ascii=False)
+        followup_input = (
+            f"你之前请求设备 {device_id} 执行了 {method} 操作，"
+            f"现在设备返回了结果：{result_str}。\n"
+            f"你之前的回复是：'{original_text}'\n"
+            f"请基于设备返回的实际结果，用自然语言重新组织回复，告诉用户结果。"
+        )
+
+        try:
+            llm_result = await self.llm_core.interact(user_input={"text": followup_input})
+            logger.info(f"[Core] 设备结果回送 LLM 完成: {llm_result}")
+        except Exception as e:
+            logger.error(f"[Core] 设备结果回送 LLM 失败: {e}")
+
+    def _on_device_message(self, device_id: str, rpc_call: str) -> Optional[str]:
+        """YosugaServer 的设备消息回调：通过 WebSocket 发送 RPC 到客户端"""
+        logger.info(f"[Core] 发送设备命令到 {device_id}")
+        asyncio.create_task(self.device_dto.send_device_command(device_id, rpc_call))
+        return None
+
+    def _on_capabilities_changed(self, capabilities: dict):
+        """设备能力变更回调：更新 LLM 系统提示词中的状态表"""
+        functions_str = json.dumps(capabilities.get("functions", []), ensure_ascii=False, indent=2)
+        device_str = json.dumps(capabilities.get("devices", {}), ensure_ascii=False, indent=2)
+        state_table = (
+            f"【当前在线设备】\n{device_str}\n\n"
+            f"【设备可用函数】\n{functions_str}"
+        )
+        self.llm_core.core_config.system_state_table = state_table
+        logger.info(f"[Core] 系统状态表已更新 | 设备: {capabilities.get('device_count', 0)} 台 | 函数: {capabilities.get('function_count', 0)} 个")
+
     def _handle_fallback(self, data: LLMCoreAnalysisBase):
         """
         llm_core同步处理器：回退处理器

src/server_core/llm_core/llm_core.py

@@ -17,6 +17,7 @@ from src.modules.text_ai_module.text_ai_core.general_text_ai_req import (
 )
 from src.server_core.llm_core.llm_core_analysis import (
     LLMCoreAnalysisManager, LLMCoreAnalysisBase, YosugaAudioResponseData, YosugaUITARSResponseData, YosugaUITARSRequestData
+    , YosugaEmbeddedResponseData
 )
 from src.server_core.llm_core.llm_core_dispatcher import LLMCoreActionDispatcher
 from src.server_core.llm_core.llm_core_prompt_manager import (

src/server_core/llm_core/llm_core_analysis.py

@@ -249,18 +249,25 @@ class YosugaLive2DResponseData(LLMCoreAnalysisBase):
 
 class YosugaEmbeddedResponseData(LLMCoreAnalysisBase):
     """
-    嵌入式设备场景的LLM输出数据模型   TODO
+    嵌入式设备场景的LLM输出数据模型
+    LLM输出 JSON-RPC 风格的函数调用，由服务端解析并路由到对应设备
     """
 
     type: str = Field(default="embedded_control", description="固定为embedded_control")
-    device_id: str = Field(..., description="设备ID")
-    command: str = Field(..., description="控制指令")
-    params: Optional[Dict[str, Any]] = Field(default=None, description="参数")
+    calls: list[dict] = Field(default_factory=list, description="JSON-RPC 调用列表，每项含 method/params/id")
+    response_text: str = Field(default="", description="同时回复给用户的文本（可选）")
 
     @classmethod
     def type_(cls) -> str:
         return "embedded_control"
 
+    def to_dict(self) -> dict:
+        return {
+            "type": self.type,
+            "calls": self.calls,
+            "response_text": self.response_text
+        }
+
 # 使用示例
 if __name__ == "__main__":
     from loguru import logger

src/server_core/llm_core/llm_core_prompt_manager.py

@@ -9,8 +9,10 @@ from pydantic import BaseModel, Field, field_validator
 from typing import Callable, List, Optional, Coroutine, Any, ClassVar, Dict
 from src.server_core.llm_core.llm_core_prompts import YOSUGA_SYSTEM_PROMPT_SCH
 
+
 class LLMCorePromptBase(BaseModel, ABC):
     """LLM 提示词基类：定义输入输出结构"""
+
     @abstractmethod
     def type(self) -> str:
         """返回该提示词类型的唯一标识"""
@@ -62,8 +64,10 @@ class LLMCorePromptManager(LLMCorePromptBase):
             for type_id, son in self._registry.items()
         )
 
+
 class YosugaAudioASRText(LLMCorePromptBase):
     """音频ASR文本输入场景"""
+
     def type(self) -> str:
         return "用户语音asr信息"
 
@@ -92,19 +96,50 @@ class YosugaAudioASRText(LLMCorePromptBase):
                 - `action`: 触发的动作指令，如"wave_hand"、"nod"等，"none"表示无动作
                 '''
 
+
 class YosugaEmbedded(LLMCorePromptBase):
-    """嵌入式设备输入场景"""
+    """嵌入式设备控制场景"""
+
     def type(self) -> str:
-        pass
+        return "嵌入式设备信息"
 
     def describe_input(self) -> str:
-        pass
+        return '''
+                当嵌入式设备有数据上报或客户端发来设备状态时，你会收到以下格式：
+                {
+                    "device_event": {
+                        "device_id": "设备ID",
+                        "event": "事件内容"
+                    }
+                }
+                但大多数情况下，你只需参考系统状态表中的设备函数表来决策。
+                '''
 
     def describe_output(self) -> str:
-        pass
+        return '''
+                当你需要控制嵌入式设备时，按以下JSON格式返回：
+                {
+                    "type": "固定为embedded_control",
+                    "calls": [
+                        {
+                            "method": "函数名（来自系统状态表中的设备能力表）",
+                            "params": { "参数名": 参数值 }
+                        }
+                    ],
+                    "response_text": "同时回复给用户的文本说明（可选，可留空）"
+                }
+                - `calls`: JSON-RPC 调用列表，每个调用对应一个设备函数
+                - `method`: 必须来自系统状态表中列出的可用函数名
+                - `params`: 按函数定义的参数传入，可省略无参数的调用
+                - `response_text`: 可选，若同时需要回复用户可在此填写
+                注意：仅当用户意图涉及现实世界控制时，才需要返回 embedded_control。
+                如果只是聊天，只需返回 audio_text。
+                '''
+
 
 class YosugaUITARS(LLMCorePromptBase):
     """自动化操作构建场景"""
+
     def type(self) -> str:
         return "自动化操作信息"
 
@@ -140,12 +175,14 @@ class YosugaUITARS(LLMCorePromptBase):
                 }
             自动化agent返回的操作信息不一定包括JSON的全部字段，例如某次返回只有key的内容，或者只有content的内容。
             针对自动化agent操作输入的返回，若没有相关内容可以留空相关字段，请不要省略掉任何字段名称。
-            
+
             注意：自动化agent的状态可见YosugaSystemState表。
         '''
 
+
 class YosugaLive2DControl(LLMCorePromptBase):
     """对Yosuga Live2D控制场景"""
+
     def type(self) -> str:
         return "Yosuga Live2D控制信息"
 
@@ -168,7 +205,7 @@ if __name__ == "__main__":
         OutputInfo=manager.describe_output(),
         RoleSetting="...",
         Language="ja",
-        Memory = "",
+        Memory="",
         SystemStateTable=""
     )
     print(system_prompt)

src/server_core/llm_core/llm_core_token.py

@@ -60,7 +60,7 @@ class TokenManager:
         )
 
     def _get_tokenizer(self, model_name: str) -> tiktoken.Encoding:
-        """获取 tokenizer（与之前实现相同，省略重复代码）"""
+        """获取 tokenizer"""
         model_tokenizer_map = {
             "qwen": "gpt-3.5-turbo",
             "llama": "gpt-3.5-turbo",
@@ -178,7 +178,7 @@ class TokenManager:
         )
 
     def count_text_tokens(self, text: str) -> int:
-        """计算单段文本的 token 数量（与之前相同）"""
+        """计算单段文本的 token 数量"""
         if not isinstance(text, str) or not text:
             return 0
         return len(self.tokenizer.encode(text))
@@ -188,7 +188,7 @@ class TokenManager:
             messages: List[Any],
             tokens_per_message: int = 3
     ) -> int:
-        """计算消息列表的总 token 数量（与之前相同，优化实现）"""
+        """计算消息列表的总 token 数量"""
         if not messages:
             return 0
 
@@ -311,7 +311,7 @@ class TokenManager:
             )
 
     def get_tokenizer_info(self) -> TokenizerInfo:
-        """获取当前 tokenizer 的详细信息（与之前相同）"""
+        """获取当前 tokenizer 的详细信息"""
         if "cl100k_base" in self.tokenizer.name and "gpt-3.5" not in self.model_name:
             accuracy = "low"
         elif self.model_name in self.tokenizer.name:
@@ -332,7 +332,7 @@ class TokenManager:
             limit: int,
             threshold: float = 0.85
     ) -> bool:
-        """判断 token 使用量是否接近限制（与之前相同）"""
+        """判断 token 使用量是否接近限制"""
         return current_tokens > limit * threshold
 
     def calculate_chunk_size(
@@ -340,7 +340,7 @@ class TokenManager:
             available_tokens: int,
             safety_margin: float = 0.1
     ) -> int:
-        """计算安全的消息块大小（与之前相同）"""
+        """计算安全的消息块大小"""
         return int(available_tokens * (1 - safety_margin))
 
     def clear_api_usage_cache(self):

src/server_core/yosuga_embedded_server/__init__.py

@@ -0,0 +1,16 @@
+"""
+Yosuga 服务端 - 面向AI的嵌入式设备JSON-RPC框架。
+
+管理多个嵌入式设备，每个设备暴露AI可调用的函数。
+服务端维护全局函数注册表、生成描述可用能力的AI提示词，
+并在AI和设备之间路由JSON-RPC调用。
+"""
+
+from .device_manager import DeviceManager, DeviceInfo, DeviceState
+from .function_registry import FunctionRegistry, FunctionInfo, ParamInfo, FuncType
+from .ai_prompt import AIPromptBuilder
+from .json_rpc import JSONRPCHandler, RPCRequest, RPCResponse, RPCError
+from .server import YosugaServer, ServerConfig
+from .device_dto import DeviceDataDTO
+
+__version__ = "0.1.0"

src/server_core/yosuga_embedded_server/ai_prompt.py

@@ -0,0 +1,169 @@
+"""
+AI提示词构建器 - 构造用于AI函数调用的结构化提示词。
+
+提示词引导AI输出有效的JSON-RPC调用，
+服务端可将这些调用转发到相应的嵌入式设备。
+"""
+
+import json
+from typing import Optional
+
+
+# -- 约束AI输出的系统提示词模板 --
+SYSTEM_PROMPT_TEMPLATE = """你是一个通过函数调用来控制嵌入式设备的AI助手。
+
+你可以访问所有已连接设备上的以下函数：
+
+{functions_table}
+
+指令:
+1. 你可以同时调用一个或多个函数。
+2. 要调用函数，请以JSON数组格式返回RPC调用：
+   {calls_syntax_example}
+3. 始终为所有必填参数提供有效值。
+4. 对于可选参数，仅在需要时包含它们。
+5. 如果函数返回数据，结果将回传给你。
+
+规则:
+- 只能调用上面列出的函数。
+- 不要虚构函数名或参数。
+- 如果需要的功能没有对应的函数支持，请说明你需要的功能。
+- 如果进行函数调用，只返回JSON数组（不要额外文字）。
+- 如果需要提问或提供信息，添加"_explanation"字段。
+
+函数调用的响应格式:
+```json
+[
+  {{
+    "jsonrpc": "2.0",
+    "method": "function_name",
+    "params": {{ "param1": value1, "param2": value2 }},
+    "id": 1
+  }}
+]
+```
+
+设备路由: 每个函数属于特定设备（显示为@device_name）。
+系统会自动将调用路由到正确的设备。
+"""
+
+
+FUNCTION_ENTRY_TEMPLATE = """### {name} (类型: {type}) @ {device_name}
+描述: {description}
+参数:
+{params}
+"""
+
+PARAM_ENTRY_TEMPLATE = "  - {name} ({type}{optional}): {description}"
+
+
+CALLS_SYNTAX_EXAMPLE = """[
+  {
+    "jsonrpc": "2.0",
+    "method": "function_name",
+    "params": {
+      "param1": "value1",
+      "param2": 42
+    },
+    "id": 1
+  }
+]"""
+
+
+class AIPromptBuilder:
+    """从函数注册表内容构建AI提示词。
+
+    提示词设计目标:
+    - 清晰列出所有可用函数及其描述
+    - 显示每个函数由哪个设备提供
+    - 约束AI输出有效的JSON-RPC
+    - 支持多调用响应
+    """
+
+    def __init__(self):
+        self._system_prompt = SYSTEM_PROMPT_TEMPLATE
+
+    def set_system_prompt(self, prompt: str):
+        self._system_prompt = prompt
+
+    def build_system_prompt(self, functions: list[dict]) -> str:
+        """用当前函数表构建完整的系统提示词。
+
+        Args:
+            functions: 来自FunctionRegistry的函数信息字典列表
+
+        Returns:
+            准备发送给AI的完整系统提示词字符串。
+        """
+        if not functions:
+            return "当前没有已连接的嵌入式设备，无可用函数。"
+
+        table = self._format_functions_table(functions)
+        return self._system_prompt.format(
+            functions_table=table,
+            calls_syntax_example=CALLS_SYNTAX_EXAMPLE,
+        )
+
+    def _format_functions_table(self, functions: list[dict]) -> str:
+        """将函数列表格式化为可读的函数表。"""
+        entries = []
+        for func in functions:
+            params_str = self._format_params(func.get("params", []))
+            entry = FUNCTION_ENTRY_TEMPLATE.format(
+                name=func["name"],
+                type=func.get("type", "ctrl_noret"),
+                device_name=func.get("device_name", "unknown"),
+                description=func.get("description", "No description"),
+                params=params_str if params_str else "  (none)",
+            )
+            entries.append(entry)
+        return "\n".join(entries)
+
+    def _format_params(self, params: list[dict]) -> str:
+        lines = []
+        for p in params:
+            optional_str = ", optional" if p.get("optional") else ""
+            lines.append(PARAM_ENTRY_TEMPLATE.format(
+                name=p.get("name", "?"),
+                type=p.get("type", "unknown"),
+                optional=optional_str,
+                description=p.get("description", ""),
+            ))
+        return "\n".join(lines)
+
+    def parse_ai_response(self, response_text: str) -> Optional[list[dict]]:
+        """从AI响应文本中提取JSON-RPC调用数组。
+
+        AI可能将JSON包裹在markdown代码块中，或直接输出原始JSON。
+        此方法处理两种情况。
+
+        Args:
+            response_text: AI返回的原始文本
+
+        Returns:
+            RPC请求字典列表，解析失败时返回None
+        """
+        text = response_text.strip()
+
+        # 尝试从markdown代码块中提取
+        if "```json" in text:
+            start = text.find("```json") + 7
+            end = text.find("```", start)
+            if end > start:
+                text = text[start:end].strip()
+        elif "```" in text:
+            start = text.find("```") + 3
+            end = text.find("```", start)
+            if end > start:
+                text = text[start:end].strip()
+
+        try:
+            data = json.loads(text)
+        except json.JSONDecodeError:
+            return None
+
+        if isinstance(data, dict):
+            return [data]
+        if isinstance(data, list):
+            return data
+        return None

src/server_core/yosuga_embedded_server/device_dto.py

@@ -0,0 +1,117 @@
+"""
+设备数据 DTO - 处理客户端发来的嵌入式设备数据
+通过 WebSocket 的 device_data 类型消息与客户端通信
+"""
+
+import json
+import time
+from typing import Callable, Optional, Coroutine
+from loguru import logger
+
+
+class DeviceDataDTO:
+    """设备数据分发器，处理客户端发来的设备注册/响应/事件"""
+
+    def __init__(self, json_dto, embedded_server):
+        """
+        Args:
+            json_dto: JsonDTO 实例，用于注册接收器
+            embedded_server: YosugaServer 实例（嵌入式框架）
+        """
+        json_dto.register_receiver("device_data", self._handle_device_data)
+        logger.info("[DeviceDataDTO] 设备数据接收业务已注册")
+        self.json_dto = json_dto
+        self.embedded_server = embedded_server
+        self._device_callbacks: list[Callable] = []
+        self.on_rpc_response: Optional[Callable[[str, dict], Coroutine]] = None
+
+    async def _handle_device_data(self, data: dict):
+        """处理设备数据的入口
+
+        客户端发来的 device_data JSON 格式:
+        {
+            "action": "register" | "rpc_response" | "event",
+            "device_id": "...",
+            "payload": { ... }         # 根据 action 不同而不同
+        }
+        """
+        action = data.get("action", "")
+        logger.info(f"[DeviceDataDTO] 收到设备数据, action={action}")
+
+        if action == "register":
+            device_id = data.get("device_id", "")
+            await self._handle_register(data.get("payload", {}), device_id)
+        elif action == "rpc_response":
+            await self._handle_rpc_response(
+                data.get("device_id", ""),
+                data.get("payload", {})
+            )
+        elif action == "event":
+            await self._handle_device_event(
+                data.get("device_id", ""),
+                data.get("payload", {})
+            )
+        else:
+            logger.warning(f"[DeviceDataDTO] 未知的 action: {action}")
+
+    async def _handle_register(self, payload: dict, device_id: str = ""):
+        """处理设备注册
+
+        payload 格式同 YosugaServer 的设备能力描述:
+        {
+            "device": { "name": "...", "description": "..." },
+            "functions": [ { "name": "...", "type": "...", ... } ]
+        }
+        """
+        try:
+            device = self.embedded_server.register_device_from_dict(payload, device_id)
+            logger.success(f"[DeviceDataDTO] 设备注册成功: {device.name} ({device.device_id})")
+        except Exception as e:
+            logger.error(f"[DeviceDataDTO] 设备注册失败: {e}")
+
+    async def _handle_rpc_response(self, device_id: str, payload: dict):
+        """处理设备返回的 RPC 响应"""
+        logger.info(f"[DeviceDataDTO] 收到设备 {device_id} 的 RPC 响应: {payload}")
+
+        if self.on_rpc_response:
+            try:
+                await self.on_rpc_response(device_id, payload)
+            except Exception as e:
+                logger.error(f"[DeviceDataDTO] on_rpc_response 回调错误: {e}")
+
+        for cb in self._device_callbacks:
+            try:
+                cb(device_id, payload)
+            except Exception as e:
+                logger.error(f"[DeviceDataDTO] 回调执行错误: {e}")
+
+    async def _handle_device_event(self, device_id: str, payload: dict):
+        """处理设备主动上报的事件"""
+        logger.info(f"[DeviceDataDTO] 收到设备 {device_id} 的事件: {payload}")
+
+    def register_device_callback(self, callback: Callable) -> None:
+        """注册设备消息回调"""
+        self._device_callbacks.append(callback)
+
+    async def send_device_command(self, device_id: str, rpc_call: str) -> None:
+        """向客户端发送设备控制命令
+
+        发送给客户端的 JSON 格式:
+        {
+            "type": "device_command",
+            "data": {
+                "device_id": "...",
+                "payload": "{\\"jsonrpc\\": \\"2.0\\", ...}"
+            }
+        }
+        """
+        payload = {
+            "device_id": device_id,
+            "payload": rpc_call
+        }
+        await self.json_dto.send_json({
+            "type": "device_command",
+            "timestamp": time.time(),
+            "data": payload
+        })
+        logger.info(f"[DeviceDataDTO] 已发送设备命令到 {device_id}")

src/server_core/yosuga_embedded_server/device_manager.py

@@ -0,0 +1,233 @@
+"""
+设备管理器 - 管理多个嵌入式设备的连接。
+
+每个设备由唯一ID标识（自动生成或客户端提供）。
+管理器处理设备注册、重复检测，并维护设备ID与其功能表的映射。
+"""
+
+import time
+import threading
+from enum import Enum
+from typing import Optional, Callable, Any
+
+
+class DeviceState(Enum):
+    """设备连接的当前状态。"""
+    DISCONNECTED = "disconnected"
+    CONNECTED = "connected"
+    REGISTERED = "registered"
+
+
+class DeviceInfo:
+    """嵌入式设备的信息。
+
+    存储设备的能力、身份和连接元数据。
+    除 device_id 外，所有字段均由设备自身提供。
+    """
+
+    def __init__(
+        self,
+        device_id: str,
+        name: str = "",
+        description: str = "",
+        firmware_version: str = "",
+        hardware_version: str = "",
+        functions: Optional[list] = None,
+    ):
+        self.device_id = device_id
+        self.name = name
+        self.description = description
+        self.firmware_version = firmware_version
+        self.hardware_version = hardware_version
+        self.functions = functions or []
+        self.state = DeviceState.DISCONNECTED
+        self.last_seen: float = 0.0
+        self.connected_at: float = 0.0
+
+    def to_dict(self) -> dict:
+        return {
+            "device_id": self.device_id,
+            "name": self.name,
+            "description": self.description,
+            "firmware_version": self.firmware_version,
+            "hardware_version": self.hardware_version,
+            "state": self.state.value,
+            "last_seen": self.last_seen,
+            "connected_at": self.connected_at,
+            "register_time": self.connected_at,
+            "functions": self.functions,
+            "function_count": len(self.functions),
+        }
+
+    def __repr__(self) -> str:
+        return f"DeviceInfo({self.device_id}, {self.name}, state={self.state.value})"
+
+
+class DeviceManager:
+    """管理所有连接到服务端的嵌入式设备。
+
+    功能:
+    - 未提供时自动生成设备ID
+    - 可配置策略处理重复设备名
+    - 跟踪连接状态和最后活跃时间戳
+    - 提供设备状态变更的回调
+    """
+
+    class ConflictStrategy(Enum):
+        """处理重复设备名的策略。"""
+        REJECT = "reject"
+        RENAME = "rename"
+        REPLACE = "replace"
+
+    def __init__(self, conflict_strategy: str = "rename"):
+        self._devices: dict[str, "DeviceInfo"] = {}
+        self._lock = threading.Lock()
+        self._next_id = 1
+        self._conflict_strategy = DeviceManager.ConflictStrategy(conflict_strategy)
+        self._on_device_change: Optional[Callable[[str, "DeviceInfo"], Any]] = None
+
+    @property
+    def on_device_change(self) -> Optional[Callable[[str, "DeviceInfo"], Any]]:
+        """设备添加、更新或移除时的回调。
+        签名: callback(event_type: str, device: DeviceInfo)
+        event_type: 'added', 'updated', 'removed'
+        """
+        return self._on_device_change
+
+    @on_device_change.setter
+    def on_device_change(self, cb: Optional[Callable[[str, "DeviceInfo"], Any]]):
+        self._on_device_change = cb
+
+    def _notify(self, event: str, device: "DeviceInfo"):
+        if self._on_device_change:
+            try:
+                self._on_device_change(event, device)
+            except Exception:
+                pass
+
+    def _generate_id(self) -> str:
+        """生成唯一的设备ID。"""
+        while True:
+            dev_id = f"device_{self._next_id}"
+            self._next_id += 1
+            if dev_id not in self._devices:
+                return dev_id
+
+    def register_from_json(self, json_data: dict) -> "DeviceInfo":
+        """从设备的能力描述JSON注册或更新设备。
+
+        期望的JSON格式:
+        {
+            "_device_id": "...",     # 可选的显式设备 ID（来自客户端转发）
+            "device": {
+                "name": "...",
+                "description": "...",
+                "firmware_version": "...",
+                "hardware_version": "..."
+            },
+            "functions": [ ... ]
+        }
+
+        返回 DeviceInfo 对象。
+        策略为 REJECT 时，发生冲突会抛出 ValueError。
+        """
+        device_data = json_data.get("device", {})
+        name = device_data.get("name", "")
+        description = device_data.get("description", "")
+
+        with self._lock:
+            explicit_id = json_data.get("_device_id", "")
+            device_id = explicit_id if explicit_id else self._resolve_identity(name)
+            device = DeviceInfo(
+                device_id=device_id,
+                name=name,
+                description=description,
+                firmware_version=device_data.get("firmware_version", ""),
+                hardware_version=device_data.get("hardware_version", ""),
+                functions=json_data.get("functions", []),
+            )
+            device.state = DeviceState.REGISTERED
+            now = time.time()
+            device.last_seen = now
+            device.connected_at = now
+
+            existing = self._find_by_name(name)
+            is_new = device_id not in self._devices
+
+            if existing and existing.device_id != device_id:
+                if self._conflict_strategy == DeviceManager.ConflictStrategy.REJECT:
+                    raise ValueError(f"Device name conflict: '{name}' already registered")
+            elif existing and existing.device_id == device_id:
+                is_new = False
+
+            self._devices[device_id] = device
+            is_new_flag = is_new
+            device_ref = device
+
+        self._notify("added" if is_new_flag else "updated", device_ref)
+        return device_ref
+
+    def _resolve_identity(self, name: str) -> str:
+        """将设备名解析为唯一的设备ID。"""
+        existing = self._find_by_name(name)
+        if existing:
+            if self._conflict_strategy == DeviceManager.ConflictStrategy.RENAME:
+                return self._generate_id()
+            elif self._conflict_strategy == DeviceManager.ConflictStrategy.REJECT:
+                raise ValueError(f"Device name conflict: '{name}' already registered")
+            elif self._conflict_strategy == DeviceManager.ConflictStrategy.REPLACE:
+                return existing.device_id
+        return self._generate_id()
+
+    def _find_by_name(self, name: str) -> Optional["DeviceInfo"]:
+        for dev in self._devices.values():
+            if dev.name == name:
+                return dev
+        return None
+
+    def get_device(self, device_id: str) -> Optional["DeviceInfo"]:
+        with self._lock:
+            return self._devices.get(device_id)
+
+    def get_device_by_name(self, name: str) -> Optional["DeviceInfo"]:
+        return self._find_by_name(name)
+
+    def remove_device(self, device_id: str) -> bool:
+        with self._lock:
+            device = self._devices.pop(device_id, None)
+            if device:
+                device.state = DeviceState.DISCONNECTED
+                self._notify("removed", device)
+                return True
+            return False
+
+    def get_all_devices(self) -> list["DeviceInfo"]:
+        with self._lock:
+            return list(self._devices.values())
+
+    def device_count(self) -> int:
+        with self._lock:
+            return len(self._devices)
+
+    def touch_device(self, device_id: str):
+        with self._lock:
+            device = self._devices.get(device_id)
+            if device:
+                device.last_seen = time.time()
+
+    def to_dict(self) -> dict:
+        return {
+            "devices": [d.to_dict() for d in self.get_all_devices()],
+            "count": self.device_count(),
+        }
+
+    def get_device_for_function(self, function_name: str) -> Optional[str]:
+        """Find which device provides a specific function name.
+        Returns device_id or None if not found.
+        """
+        with self._lock:
+            for dev_id, dev in self._devices.items():
+                for func in dev.functions:
+                    if func.get("name") == function_name:
+                        return dev_id
+            return None

src/server_core/yosuga_embedded_server/function_registry.py

@@ -0,0 +1,190 @@
+"""
+函数注册表 - 维护所有设备的全局函数表。
+
+这是所有已连接设备的所有函数的汇总视图。
+注册表在运行时动态更新：随设备连接/断开而更新。
+"""
+
+import copy
+import json
+import threading
+from enum import Enum
+from typing import Optional, Callable
+
+
+class FuncType(Enum):
+    CTRL_NORET = "ctrl_noret"
+    CTRL_RET = "ctrl_ret"
+    DATA_RET = "data_ret"
+    NORET_NODATA = "noret_nodata"
+
+    @classmethod
+    def from_str(cls, s: str) -> "FuncType":
+        for ft in cls:
+            if ft.value == s:
+                return ft
+        return cls.CTRL_NORET
+
+
+class ParamInfo:
+    """单个函数参数的描述符"""
+
+    def __init__(self, name: str = "", description: str = "",
+                 param_type: str = "int", optional: bool = False):
+        self.name = name
+        self.description = description
+        self.type = param_type
+        self.optional = optional
+
+    @classmethod
+    def from_dict(cls, d: dict) -> "ParamInfo":
+        return cls(
+            name=d.get("name", ""),
+            description=d.get("description", ""),
+            param_type=d.get("type", "int"),
+            optional=d.get("optional", False),
+        )
+
+    def to_dict(self) -> dict:
+        return {
+            "name": self.name,
+            "description": self.description,
+            "type": self.type,
+            "optional": self.optional,
+        }
+
+    def __repr__(self) -> str:
+        return f"ParamInfo({self.name}: {self.type})"
+
+
+class FunctionInfo:
+    """整个系统中单个函数的完整描述符。
+
+    包含函数名、描述、参数信息、类型以及提供该函数的设备。
+    """
+
+    def __init__(
+        self,
+        name: str = "",
+        description: str = "",
+        func_type: FuncType = FuncType.CTRL_NORET,
+        params: Optional[list] = None,
+        device_id: str = "",
+        device_name: str = "",
+    ):
+        self.name = name
+        self.description = description
+        self.func_type = func_type
+        self.params = params or []
+        self.device_id = device_id
+        self.device_name = device_name
+
+    @classmethod
+    def from_device_func(cls, func_dict: dict, device_id: str, device_name: str) -> "FunctionInfo":
+        raw_params = func_dict.get("params", [])
+        params = [ParamInfo.from_dict(p) if isinstance(p, dict) else ParamInfo()
+                  for p in raw_params]
+        return cls(
+            name=func_dict.get("name", ""),
+            description=func_dict.get("description", ""),
+            func_type=FuncType.from_str(func_dict.get("type", "ctrl_noret")),
+            params=params,
+            device_id=device_id,
+            device_name=device_name,
+        )
+
+    def to_dict(self) -> dict:
+        return {
+            "name": self.name,
+            "description": self.description,
+            "type": self.func_type.value,
+            "params": [p.to_dict() for p in self.params],
+            "device_id": self.device_id,
+            "device_name": self.device_name,
+        }
+
+    def __repr__(self) -> str:
+        return f"FunctionInfo({self.name} @ {self.device_name})"
+
+
+class FunctionRegistry:
+    """所有设备上所有函数的全局注册表。
+
+    维护:
+    - functions_by_name: dict[str, FunctionInfo] - 按名称快速查找
+    - functions_by_device: dict[str, list[FunctionInfo]] - 按设备查找
+
+    线程安全。变更时触发回调。
+    """
+
+    def __init__(self):
+        self._lock = threading.Lock()
+        self._functions_by_name: dict[str, FunctionInfo] = {}
+        self._functions_by_device: dict[str, list[FunctionInfo]] = {}
+        self._on_change: Optional[Callable] = None
+
+    @property
+    def on_change(self) -> Optional[Callable]:
+        return self._on_change
+
+    @on_change.setter
+    def on_change(self, cb: Optional[Callable]):
+        self._on_change = cb
+
+    def _notify(self):
+        if self._on_change:
+            try:
+                self._on_change()
+            except Exception:
+                pass
+
+    def add_device_functions(self, device_id: str, device_name: str,
+                             func_list: list[dict]):
+        """添加或更新某个设备的所有函数。"""
+        with self._lock:
+            self._remove_device_functions_locked(device_id)
+            func_infos = []
+            for func_dict in func_list:
+                fi = FunctionInfo.from_device_func(func_dict, device_id, device_name)
+                func_infos.append(fi)
+                self._functions_by_name[fi.name] = fi
+            self._functions_by_device[device_id] = func_infos
+        self._notify()
+
+    def _remove_device_functions_locked(self, device_id: str):
+        """移除设备函数，不加锁（调用者必须持有锁）。"""
+        funcs = self._functions_by_device.pop(device_id, [])
+        for fi in funcs:
+            self._functions_by_name.pop(fi.name, None)
+
+    def remove_device_functions(self, device_id: str):
+        """移除某个设备的所有函数。"""
+        with self._lock:
+            self._remove_device_functions_locked(device_id)
+        self._notify()
+
+    def get_function(self, name: str) -> Optional[FunctionInfo]:
+        with self._lock:
+            return self._functions_by_name.get(name)
+
+    def get_all_functions(self) -> list[FunctionInfo]:
+        with self._lock:
+            return list(self._functions_by_name.values())
+
+    def get_device_functions(self, device_id: str) -> list[FunctionInfo]:
+        with self._lock:
+            return list(self._functions_by_device.get(device_id, []))
+
+    def function_count(self) -> int:
+        with self._lock:
+            return len(self._functions_by_name)
+
+    def to_function_list(self) -> list[dict]:
+        return [fi.to_dict() for fi in self.get_all_functions()]
+
+    def to_json(self) -> str:
+        return json.dumps(self.to_function_list(), indent=2, ensure_ascii=False)
+
+    def find_device_for_function(self, func_name: str) -> Optional[str]:
+        fi = self.get_function(func_name)
+        return fi.device_id if fi else None

src/server_core/yosuga_embedded_server/json_rpc.py

@@ -0,0 +1,232 @@
+"""
+JSON-RPC 2.0 协议处理器。
+
+处理AI、服务端和嵌入式设备之间JSON-RPC消息的
+解析、验证、构建和路由
+"""
+
+import json
+from typing import Optional
+
+
+class RPCError(Exception):
+    """JSON-RPC错误，包含标准错误码"""
+
+    # 标准 JSON-RPC 错误码
+    PARSE_ERROR = -32700
+    INVALID_REQUEST = -32600
+    METHOD_NOT_FOUND = -32601
+    INVALID_PARAMS = -32602
+    INTERNAL_ERROR = -32603
+
+    # 自定义错误码
+    DEVICE_NOT_FOUND = -32000
+    DEVICE_ERROR = -32001
+    TIMEOUT = -32002
+
+    def __init__(self, code: int, message: str, data: Optional[dict] = None):
+        self.code = code
+        self.message = message
+        self.data = data
+        super().__init__(f"[{code}] {message}")
+
+    def to_dict(self) -> dict:
+        err = {"code": self.code, "message": self.message}
+        if self.data:
+            err["data"] = self.data
+        return err
+
+
+class RPCRequest:
+    """表示一个JSON-RPC 2.0请求"""
+
+    def __init__(self, method: str, params: Optional[dict] = None,
+                 request_id: Optional[int] = None):
+        self.method = method
+        self.params = params or {}
+        self.id = request_id
+
+    def is_notification(self) -> bool:
+        return self.id is None
+
+    def to_dict(self) -> dict:
+        req = {
+            "jsonrpc": "2.0",
+            "method": self.method,
+        }
+        if self.params:
+            req["params"] = self.params
+        if self.id is not None:
+            req["id"] = self.id
+        return req
+
+    def to_json(self) -> str:
+        return json.dumps(self.to_dict(), ensure_ascii=False)
+
+    @classmethod
+    def from_dict(cls, d: dict) -> "RPCRequest":
+        return cls(
+            method=d["method"],
+            params=d.get("params"),
+            request_id=d.get("id"),
+        )
+
+    def __repr__(self) -> str:
+        return f"RPCRequest(method={self.method}, id={self.id})"
+
+
+class RPCResponse:
+    """表示一个JSON-RPC 2.0响应"""
+
+    def __init__(self, result: Optional[dict] = None,
+                 error: Optional[RPCError] = None,
+                 request_id: Optional[int] = None):
+        self.result = result
+        self.error = error
+        self.id = request_id
+
+    def is_success(self) -> bool:
+        return self.error is None
+
+    def to_dict(self) -> dict:
+        resp = {"jsonrpc": "2.0"}
+        if self.id is not None:
+            resp["id"] = self.id
+        if self.error:
+            resp["error"] = self.error.to_dict()
+        else:
+            resp["result"] = self.result
+        return resp
+
+    def to_json(self) -> str:
+        return json.dumps(self.to_dict(), ensure_ascii=False)
+
+    @classmethod
+    def success(cls, result: Optional[dict], request_id: Optional[int]) -> "RPCResponse":
+        return cls(result=result, request_id=request_id)
+
+    @classmethod
+    def error(cls, code: int, message: str, request_id: Optional[int] = None) -> "RPCResponse":
+        return cls(error=RPCError(code, message), request_id=request_id)
+
+    def __repr__(self) -> str:
+        if self.error:
+            return f"RPCResponse(error={self.error.message}, id={self.id})"
+        return f"RPCResponse(result={self.result}, id={self.id})"
+
+
+class JSONRPCHandler:
+    """处理JSON-RPC协议解析和响应构建"""
+
+    @staticmethod
+    def parse_request(json_str: str) -> Optional[RPCRequest]:
+        """将JSON字符串解析为RPCRequest"""
+        try:
+            data = json.loads(json_str)
+        except json.JSONDecodeError:
+            return None
+
+        if not isinstance(data, dict):
+            return None
+        if data.get("jsonrpc") != "2.0":
+            return None
+        if "method" not in data or not isinstance(data["method"], str):
+            return None
+
+        return RPCRequest.from_dict(data)
+
+    @staticmethod
+    def parse_request_batch(json_str: str) -> Optional[list[RPCRequest]]:
+        """将JSON字符串解析为RPCRequest列表（用于批量调用）"""
+        try:
+            data = json.loads(json_str)
+        except json.JSONDecodeError:
+            return None
+
+        if isinstance(data, dict):
+            req = JSONRPCHandler.parse_request(json_str)
+            return [req] if req else None
+
+        if isinstance(data, list):
+            results = []
+            for item in data:
+                item_str = json.dumps(item)
+                req = JSONRPCHandler.parse_request(item_str)
+                if req:
+                    results.append(req)
+            return results if results else None
+
+        return None
+
+    @staticmethod
+    def validate_request(request: dict) -> Optional[RPCError]:
+        """验证原始请求字典。无效时返回RPCError"""
+        if not isinstance(request, dict):
+            return RPCError(RPCError.INVALID_REQUEST, "request must be a JSON object")
+        if request.get("jsonrpc") != "2.0":
+            return RPCError(RPCError.INVALID_REQUEST, "jsonrpc must be '2.0'")
+        if "method" not in request:
+            return RPCError(RPCError.INVALID_REQUEST, "missing method")
+        if not isinstance(request["method"], str) or not request["method"]:
+            return RPCError(RPCError.INVALID_REQUEST, "method must be a non-empty string")
+        params = request.get("params")
+        if params is not None and not isinstance(params, dict):
+            return RPCError(RPCError.INVALID_PARAMS, "params must be a JSON object")
+        return None
+
+    @staticmethod
+    def build_error_response(code: int, message: str,
+                             request_id: Optional[int] = None) -> str:
+        """构建JSON-RPC错误响应字符串"""
+        resp = RPCResponse.error(code, message, request_id)
+        return resp.to_json()
+
+    @staticmethod
+    def build_success_response(result: Optional[dict],
+                               request_id: Optional[int] = None) -> str:
+        """构建JSON-RPC成功响应字符串"""
+        resp = RPCResponse.success(result, request_id)
+        return resp.to_json()
+
+    @staticmethod
+    def build_call(method: str, params: Optional[dict] = None,
+                   call_id: Optional[int] = None) -> str:
+        """构建JSON-RPC调用字符串（服务端 -> 设备）"""
+        req = RPCRequest(method, params, call_id)
+        return req.to_json()
+
+    @staticmethod
+    def is_response(json_str: str) -> bool:
+        """检查JSON字符串是否为JSON-RPC响应"""
+        try:
+            data = json.loads(json_str)
+            if isinstance(data, dict):
+                return "result" in data or "error" in data
+            return False
+        except json.JSONDecodeError:
+            return False
+
+    @staticmethod
+    def parse_response(json_str: str) -> Optional[RPCResponse]:
+        """将JSON字符串解析为RPCResponse"""
+        try:
+            data = json.loads(json_str)
+        except json.JSONDecodeError:
+            return None
+
+        if not isinstance(data, dict):
+            return None
+        if data.get("jsonrpc") != "2.0":
+            return None
+
+        request_id = data.get("id")
+        if "error" in data:
+            err_data = data["error"]
+            return RPCResponse.error(
+                err_data.get("code", RPCError.INTERNAL_ERROR),
+                err_data.get("message", "unknown error"),
+                request_id,
+            )
+        elif "result" in data:
+            return RPCResponse.success(data["result"], request_id)
+        return None

src/server_core/yosuga_embedded_server/server.py

@@ -0,0 +1,270 @@
+"""
+YosugaServer - 主服务端类，串联所有组件。
+
+架构:
+  AI <-> YosugaServer <-> DeviceManager <-> 嵌入式设备
+                           FunctionRegistry
+                           AIPromptBuilder
+                           JSONRPCHandler
+
+流程:
+  1. 设备连接，发送能力描述JSON
+  2. 服务端注册设备+函数
+  3. 变更回调触发 -> 服务端更新AI提示词
+  4. 用户通过服务端向AI发送请求
+  5. 服务端用当前函数构建系统提示词
+  6. AI响应函数调用
+  7. 服务端解析调用并路由到设备
+  8. 设备响应被收集并返回
+"""
+
+import json
+import logging
+import threading
+from typing import Optional, Callable, Any
+
+from .device_manager import DeviceManager, DeviceInfo
+from .function_registry import FunctionRegistry
+from .ai_prompt import AIPromptBuilder
+from .json_rpc import JSONRPCHandler, RPCRequest, RPCResponse, RPCError
+
+logger = logging.getLogger(__name__)
+
+
+class ServerConfig:
+    """服务端配置。"""
+
+    def __init__(
+        self,
+        device_conflict_strategy: str = "rename",
+        max_concurrent_calls: int = 10,
+        device_timeout: float = 30.0,
+    ):
+        self.device_conflict_strategy = device_conflict_strategy
+        self.max_concurrent_calls = max_concurrent_calls
+        self.device_timeout = device_timeout
+
+
+class YosugaServer:
+    """主服务端，协调设备、函数和AI交互。"""
+
+    def __init__(self, config: Optional[ServerConfig] = None):
+        self.config = config or ServerConfig()
+        self.device_manager = DeviceManager(
+            conflict_strategy=self.config.device_conflict_strategy
+        )
+        self.function_registry = FunctionRegistry()
+        self.ai_prompt = AIPromptBuilder()
+        self._lock = threading.Lock()
+        self._call_id_counter = 0
+
+        # 挂载变更通知
+        self.device_manager.on_device_change = self._on_device_change
+        self.function_registry.on_change = self._on_functions_change
+
+        # 外部集成回调
+        self.on_capabilities_changed: Optional[Callable[[dict], Any]] = None
+        self.on_device_message: Optional[Callable[[str, str], Optional[str]]] = None
+
+    def _on_device_change(self, event: str, device: DeviceInfo):
+        logger.info("Device %s: %s", event, device.device_id)
+        if event == "removed":
+            self.function_registry.remove_device_functions(device.device_id)
+        elif event in ("added", "updated"):
+            if device.state.value == "registered":
+                self.function_registry.add_device_functions(
+                    device.device_id,
+                    device.name,
+                    device.functions or [],
+                )
+
+    def _on_functions_change(self):
+        if self.on_capabilities_changed:
+            try:
+                self.on_capabilities_changed(self.get_capabilities_summary())
+            except Exception as e:
+                logger.error("capabilities callback error: %s", e)
+
+    def _next_call_id(self) -> int:
+        with self._lock:
+            self._call_id_counter += 1
+            return self._call_id_counter
+
+    def register_device(self, device_json_str: str) -> DeviceInfo:
+        """从设备的JSON能力描述注册设备。
+
+        Args:
+            device_json_str: 设备发来的JSON字符串
+
+        Returns:
+            已注册设备的 DeviceInfo
+        """
+        data = json.loads(device_json_str)
+        return self.device_manager.register_from_json(data)
+
+    def register_device_from_dict(self, device_json: dict, device_id: str = "") -> DeviceInfo:
+        """从字典注册设备（格式同JSON）。
+        Args:
+            device_json: 设备能力描述字典
+            device_id: 可选的显式设备 ID（来自客户端转发，覆盖自动生成）
+        """
+        if device_id:
+            device_json = dict(device_json)
+            device_json["_device_id"] = device_id
+        return self.device_manager.register_from_json(device_json)
+
+    def remove_device(self, device_id: str) -> bool:
+        """移除设备及其函数。"""
+        return self.device_manager.remove_device(device_id)
+
+    def build_ai_system_prompt(self) -> str:
+        """构建当前AI系统提示词。"""
+        functions = self.function_registry.to_function_list()
+        return self.ai_prompt.build_system_prompt(functions)
+
+    def process_ai_response(self, response_text: str) -> list[dict]:
+        """解析AI响应为RPC调用并路由到设备。
+
+        Args:
+            response_text: AI返回的原始文本
+
+        Returns:
+            包含设备响应结果的字典列表
+        """
+        calls = self.ai_prompt.parse_ai_response(response_text)
+        if not calls:
+            return [{"error": "无法将AI响应解析为RPC调用"}]
+
+        results = []
+        for call in calls:
+            method = call.get("method")
+            params = call.get("params", {})
+            call_id = call.get("id", self._next_call_id())
+
+            # 查找哪个设备提供此函数
+            func_info = self.function_registry.get_function(method)
+            if not func_info:
+                results.append({
+                    "id": call_id,
+                    "method": method,
+                    "error": {"code": RPCError.METHOD_NOT_FOUND,
+                              "message": f"未找到函数 '{method}'"},
+                })
+                continue
+
+            device_id = func_info.device_id
+            device = self.device_manager.get_device(device_id)
+            if not device:
+                results.append({
+                    "id": call_id,
+                    "method": method,
+                    "error": {"code": RPCError.DEVICE_NOT_FOUND,
+                              "message": f"设备 '{device_id}' 不可用"},
+                })
+                continue
+
+            # 构建发送给设备的RPC调用
+            rpc_call = JSONRPCHandler.build_call(method, params, call_id)
+
+            # 如果有设备消息回调，使用它
+            if self.on_device_message:
+                try:
+                    response_str = self.on_device_message(device_id, rpc_call)
+                    if response_str:
+                        resp = JSONRPCHandler.parse_response(response_str)
+                        if resp:
+                            if resp.is_success():
+                                results.append({
+                                    "id": call_id,
+                                    "method": method,
+                                    "device_id": device_id,
+                                    "result": resp.result,
+                                })
+                            else:
+                                results.append({
+                                    "id": call_id,
+                                    "method": method,
+                                    "device_id": device_id,
+                                    "error": resp.error.to_dict(),
+                                })
+                        else:
+                            results.append({
+                                "id": call_id,
+                                "method": method,
+                                "error": {"code": RPCError.PARSE_ERROR,
+                                          "message": "Invalid response from device"},
+                            })
+                    else:
+                        results.append({
+                            "id": call_id,
+                            "method": method,
+                            "device_id": device_id,
+                            "result": None,
+                            "note": "notification (no response expected)",
+                        })
+                except Exception as e:
+                    results.append({
+                        "id": call_id,
+                        "method": method,
+                        "error": {"code": RPCError.DEVICE_ERROR,
+                                  "message": str(e)},
+                    })
+            else:
+                results.append({
+                    "id": call_id,
+                    "method": method,
+                    "device_id": device_id,
+                    "note": "No on_device_message callback set - call would be routed here",
+                })
+
+        return results
+
+    def list_devices(self) -> list[dict]:
+        """获取所有在线设备的字典列表"""
+        return [d.to_dict() for d in self.device_manager.get_all_devices()]
+
+    def send_rpc(self, device_id: str, rpc_call: str) -> Optional[str]:
+        """向指定设备发送 RPC 调用并返回响应"""
+        if self.on_device_message:
+            return self.on_device_message(device_id, rpc_call)
+        return None
+
+    def get_capabilities_summary(self) -> dict:
+        """获取所有能力的摘要（供外部使用）。"""
+        return {
+            "device_count": self.device_manager.device_count(),
+            "function_count": self.function_registry.function_count(),
+            "devices": self.device_manager.to_dict(),
+            "functions": self.function_registry.to_function_list(),
+        }
+
+    def process_device_message(self, device_id: str, message: str) -> str:
+        """处理设备发来的消息。
+
+        设备发送:
+        - 能力描述（注册）
+        - RPC响应
+
+        如有需要返回响应字符串。
+        """
+        try:
+            data = json.loads(message)
+        except json.JSONDecodeError:
+            return JSONRPCHandler.build_error_response(
+                RPCError.PARSE_ERROR, "Invalid JSON"
+            )
+
+        # 检查是否为能力描述（包含"device"和"functions"）
+        if isinstance(data, dict) and "device" in data and "functions" in data:
+            device = self.register_device_from_dict(data)
+            return json.dumps({"status": "registered", "device_id": device.device_id})
+
+        # 检查是否为RPC响应
+        if isinstance(data, dict) and ("result" in data or "error" in data):
+            # 仅确认 - 响应由回调处理
+            return json.dumps({"status": "received"})
+
+        # 可能是转发调用或其他消息
+        return JSONRPCHandler.build_error_response(
+            RPCError.INVALID_REQUEST, "Unknown message type"
+        )

src/server_view/backend/.gitignore

@@ -0,0 +1,17 @@
+# Python-generated files
+__pycache__/
+*.py[oc]
+build/
+dist/
+wheels/
+*.egg-info
+
+# from ide
+.vscode
+.idea
+
+# Virtual environments
+.venv
+venv
+
+user_preferences.json

src/server_view/backend/app.py

@@ -0,0 +1,686 @@
+"""
+Yosuga Server Web UI - FastAPI Backend with Socket.IO
+"""
+import asyncio
+import json
+import logging
+import sys
+import time
+import psutil
+from contextlib import asynccontextmanager
+from pathlib import Path
+from typing import Dict, Any, Set
+
+from fastapi import FastAPI, HTTPException, UploadFile, File
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi.responses import JSONResponse
+from fastapi.staticfiles import StaticFiles
+import socketio
+
+# 项目根目录
+project_root = Path(__file__).parent.parent.parent
+sys.path.insert(0, str(project_root))
+
+from src.server_view.backend.core_manager import (
+    start_core, stop_core, get_status, get_core_status
+)
+from src.server_view.backend.diagnostics import get_diagnostics, HealthStatus, CheckResult
+
+# Socket.IO 服务器
+sio = socketio.AsyncServer(
+    cors_allowed_origins="*",
+    async_mode="asgi",
+    logger=False,
+    engineio_logger=False
+)
+
+# 跟踪已连接的客户端
+connected_clients: Set[str] = set()
+
+# RPC 响应转发回调标志（只注册一次）
+_rpc_forwarder_registered: bool = False
+
+# 日志广播处理器
+class SocketIOLogHandler(logging.Handler):
+    """将Python标准日志发送到Socket.IO"""
+    def __init__(self, sio_server):
+        super().__init__()
+        self.sio = sio_server
+        self.setLevel(logging.DEBUG)
+
+    def emit(self, record: logging.LogRecord):
+        try:
+            msg = self.format(record)
+            try:
+                loop = asyncio.get_running_loop()
+                loop.create_task(self._broadcast(msg, record.levelname))
+            except RuntimeError:
+                pass
+        except Exception:
+            pass
+
+    async def _broadcast(self, message: str, level: str):
+        try:
+            await self.sio.emit('log_line', {
+                'line': message,
+                'timestamp': time.time(),
+                'level': level
+            })
+        except Exception:
+            pass
+
+def setup_logging(sio_server):
+    """配置日志系统 - 减少HTTP访问日志噪音"""
+    root_logger = logging.getLogger()
+    root_logger.setLevel(logging.DEBUG)
+
+    # 清除现有处理器
+    for handler in root_logger.handlers[:]:
+        root_logger.removeHandler(handler)
+
+    # 控制台处理器 - 过滤掉频繁的系统信息HTTP请求日志
+    console = logging.StreamHandler(sys.stdout)
+    console.setLevel(logging.INFO)
+    console.setFormatter(logging.Formatter(
+        '%(asctime)s | %(levelname)-8s | %(name)s:%(lineno)d - %(message)s'
+    ))
+
+    # 添加过滤器，排除频繁的HTTP轮询日志
+    def filter_http_poll(record):
+        msg = record.getMessage()
+        # 过滤掉 /api/system/info 和 /api/core/status 的GET请求日志
+        if 'GET /api/system/info' in msg or 'GET /api/core/status' in msg:
+            return False
+        return True
+
+    console.addFilter(filter_http_poll)
+    root_logger.addHandler(console)
+
+    # Socket.IO处理器
+    sio_handler = SocketIOLogHandler(sio_server)
+    sio_handler.setLevel(logging.DEBUG)
+    sio_handler.setFormatter(logging.Formatter(
+        '%(asctime)s | %(levelname)-8s | %(name)s - %(message)s'
+    ))
+    root_logger.addHandler(sio_handler)
+
+    # 设置yosuga logger
+    yosuga_logger = logging.getLogger("yosuga")
+    yosuga_logger.setLevel(logging.DEBUG)
+
+    return root_logger
+
+# 系统状态监控
+async def system_monitor_task():
+    """后台任务：定期采集并广播系统状态"""
+    while True:
+        try:
+            # 只有有客户端连接时才采集数据
+            if connected_clients:
+                # 采集系统数据
+                cpu = psutil.cpu_percent(interval=0.1)
+                mem = psutil.virtual_memory()
+                disk = psutil.disk_usage('/')
+                proc = psutil.Process()
+
+                system_data = {
+                    "cpu": {"percent": cpu, "count": psutil.cpu_count()},
+                    "memory": {
+                        "total": mem.total, "available": mem.available,
+                        "percent": mem.percent, "used": mem.used, "free": mem.free
+                    },
+                    "disk": {
+                        "total": disk.total, "used": disk.used,
+                        "free": disk.free, "percent": (disk.used/disk.total)*100
+                    },
+                    "process": {
+                        "memory_percent": proc.memory_percent(),
+                        "cpu_percent": proc.cpu_percent(interval=0.1),
+                        "threads": proc.num_threads(),
+                        "uptime": time.time() - proc.create_time()
+                    },
+                    "timestamp": time.time()
+                }
+
+                # 广播给所有客户端
+                await sio.emit('system_stats', {
+                    'success': True,
+                    'data': system_data
+                })
+
+                # 同时广播核心状态（实时推送）
+                core_status = get_core_status()
+                await sio.emit('core_status', {
+                    'success': True,
+                    'data': core_status
+                })
+
+        except Exception as e:
+            logging.error(f"系统监控任务异常: {e}")
+
+        # 1秒间隔，比HTTP轮询更实时
+        await asyncio.sleep(1)
+
+# FastAPI 应用
+@asynccontextmanager
+async def lifespan(app: FastAPI):
+    """应用生命周期"""
+    logger = setup_logging(sio)
+    logger.info("Yosuga Server Web UI 启动")
+
+    monitor_task = asyncio.create_task(system_monitor_task())
+    logger.info("系统监控任务已启动")
+
+    yield
+
+    monitor_task.cancel()
+    try:
+        await monitor_task
+    except asyncio.CancelledError:
+        pass
+
+    # 清理诊断模块
+    try:
+        diag = await get_diagnostics()
+        # 诊断模块无需要关闭的资源，但保留钩子
+    except:
+        pass
+
+    logger.info("应用关闭")
+    stop_core()
+fastapi_app = FastAPI(
+    title="Yosuga Server Web UI",
+    version="1.0.0",
+    lifespan=lifespan
+)
+
+fastapi_app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+
+@fastapi_app.post("/api/diagnostics/run")
+async def run_diagnostics():
+    """执行完整系统体检"""
+    try:
+        diag = await get_diagnostics()
+        report = await diag.run_full_diagnostics()
+        return report.to_dict()
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+
+@fastapi_app.get("/api/diagnostics/check/{module}")
+async def check_single_module(module: str):
+    """检查单个模块状态"""
+    try:
+        from src.config.config import cfg
+        diag = await get_diagnostics()
+
+        # 获取对应配置
+        config_map = {
+            'asr': cfg.asr,
+            'tts': cfg.tts,
+            'ai': cfg.ai,
+            'auto_agent': cfg.auto_agent,
+            'llm_core': cfg.llm_core
+        }
+
+        if module not in config_map:
+            raise HTTPException(status_code=400, detail=f"未知模块: {module}")
+
+        # 转换为dict
+        config_dict = {}
+        if hasattr(config_map[module], '__dataclass_fields__'):
+            from dataclasses import asdict
+            config_dict = asdict(config_map[module])
+        else:
+            config_dict = dict(config_map[module])
+
+        result = await diag.quick_check_module(module, config_dict)
+        return {"success": True, "data": result.to_dict()}
+
+    except HTTPException:
+        raise
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+
+@fastapi_app.get("/api/diagnostics/health")
+async def quick_health_check():
+    """快速健康检查（用于负载均衡/心跳）"""
+    try:
+        from src.server_view.backend.core_manager import get_status
+        status = get_status()
+
+        # 简单检查WebSocket服务器是否活着
+        ws_ok = len(connected_clients) >= 0  # 总是True，只要能响应
+
+        health = {
+            "status": "healthy" if status.is_running else "degraded",
+            "web_ui": "up",
+            "core_running": status.is_running,
+            "websocket_clients": len(connected_clients),
+            "timestamp": time.time()
+        }
+
+        return health
+
+    except Exception as e:
+        return {"status": "unhealthy", "error": str(e)}
+
+# Socket.IO ASGI应用
+app = socketio.ASGIApp(sio, fastapi_app)
+
+# Socket.IO 事件
+@sio.event
+async def connect(sid, environ):
+    """客户端连接"""
+    connected_clients.add(sid)
+    print(f"客户端连接: {sid} (当前在线: {len(connected_clients)})")
+
+    # 立即发送一次当前状态（避免前端等待）
+    await sio.emit('system', {
+        'message': '连接成功',
+        'timestamp': time.time(),
+        'clients_count': len(connected_clients)
+    }, to=sid)
+
+    # 立即推送一次系统状态（前端无需再发HTTP请求）
+    try:
+        core_status = get_core_status()
+        await sio.emit('core_status', {
+            'success': True,
+            'data': core_status
+        }, to=sid)
+    except Exception as e:
+        logging.error(f"推送初始状态失败: {e}")
+
+@sio.event
+async def disconnect(sid):
+    """客户端断开"""
+    connected_clients.discard(sid)
+    print(f"客户端断开: {sid} (当前在线: {len(connected_clients)})")
+
+@sio.on('subscribe_logs')
+async def handle_subscribe_logs(sid, data):
+    """订阅日志（可指定级别过滤）"""
+    level = data.get('level', 'ALL') if isinstance(data, dict) else 'ALL'
+    print(f"客户端 {sid} 订阅日志: {level}")
+    await sio.emit('system', {'message': f'已订阅日志: {level}'}, to=sid)
+
+@sio.on('control_core')
+async def handle_control_core(sid, data):
+    """WebSocket方式控制核心"""
+    action = data.get('action') if isinstance(data, dict) else None
+
+    if action == 'start':
+        try:
+            # 检查是否已运行
+            status = get_status()
+            if status.is_running:
+                await sio.emit('core_control_result', {
+                    'success': True,
+                    'message': '核心已在运行',
+                    'data': status.to_dict()
+                }, to=sid)
+                return
+
+            # 启动核心
+            success, error = start_core(project_root)
+            if success:
+                # 等待一下确保启动成功
+                await asyncio.sleep(0.5)
+                new_status = get_status()
+                # 广播给所有客户端（不仅仅是操作者）
+                await sio.emit('core_status', {
+                    'success': True,
+                    'data': new_status.to_dict(),
+                    'message': '核心启动成功'
+                })
+            else:
+                await sio.emit('core_control_result', {
+                    'success': False,
+                    'error': error or '启动失败'
+                }, to=sid)
+        except Exception as e:
+            await sio.emit('core_control_result', {
+                'success': False,
+                'error': str(e)
+            }, to=sid)
+
+    elif action == 'stop':
+        try:
+            status = get_status()
+            if not status.is_running:
+                await sio.emit('core_control_result', {
+                    'success': True,
+                    'message': '核心已停止',
+                    'data': status.to_dict()
+                }, to=sid)
+                return
+
+            success, error = stop_core()
+            if success:
+                await asyncio.sleep(0.5)  # 等待停止完成
+                new_status = get_status()
+                await sio.emit('core_status', {
+                    'success': True,
+                    'data': new_status.to_dict(),
+                    'message': '核心已停止'
+                })
+            else:
+                await sio.emit('core_control_result', {
+                    'success': False,
+                    'error': error or '停止失败'
+                }, to=sid)
+        except Exception as e:
+            await sio.emit('core_control_result', {
+                'success': False,
+                'error': str(e)
+            }, to=sid)
+
+
+# 设备管理 Socket.IO 事件
+@sio.on('get_devices')
+async def handle_get_devices(sid):
+    """获取当前所有在线设备列表"""
+    try:
+        from src.server_core.core import YosugaServerCore
+        core = await YosugaServerCore.get_instance()
+        devices = core.embedded_server.list_devices()
+        await sio.emit('devices_list', {
+            'success': True,
+            'data': devices
+        }, to=sid)
+    except Exception as e:
+        await sio.emit('devices_list', {
+            'success': False,
+            'error': str(e)
+        }, to=sid)
+
+@sio.on('send_device_rpc')
+async def handle_send_device_rpc(sid, data):
+    """向指定设备发送 RPC 命令"""
+    device_id = data.get('device_id') if isinstance(data, dict) else None
+    rpc_call = data.get('rpc_call') if isinstance(data, dict) else None
+    if not device_id or not rpc_call:
+        await sio.emit('device_rpc_result', {
+            'success': False,
+            'error': '缺少 device_id 或 rpc_call'
+        }, to=sid)
+        return
+    try:
+        from src.server_core.core import YosugaServerCore
+        core = await YosugaServerCore.get_instance()
+
+        # 注册一次性的 RPC 响应转发回调
+        global _rpc_forwarder_registered
+        if not _rpc_forwarder_registered:
+            async def forward_rpc_response(dev_id: str, payload: dict):
+                await sio.emit('device_rpc_response', {
+                    'device_id': dev_id,
+                    'payload': payload
+                })
+            core.device_dto.on_rpc_response = forward_rpc_response
+            _rpc_forwarder_registered = True
+
+        core.embedded_server.send_rpc(device_id, rpc_call)
+        await sio.emit('device_rpc_result', {
+            'success': True,
+            'device_id': device_id,
+            'message': 'RPC 命令已发送到设备'
+        }, to=sid)
+    except Exception as e:
+        await sio.emit('device_rpc_result', {
+            'success': False,
+            'error': str(e)
+        }, to=sid)
+
+# 设备管理 REST API
+@fastapi_app.get("/api/devices")
+async def get_devices_api():
+    """获取所有在线设备（HTTP 备用）"""
+    try:
+        from src.server_core.core import YosugaServerCore
+        core = await YosugaServerCore.get_instance()
+        devices = core.embedded_server.list_devices()
+        return {"success": True, "data": devices}
+    except Exception as e:
+        return {"success": False, "error": str(e)}
+
+@sio.on('check_module_health')
+async def handle_check_module(sid, data):
+    """WebSocket方式检查模块健康"""
+    module = data.get('module') if isinstance(data, dict) else None
+    if not module:
+        await sio.emit('module_health_result', {
+            "success": False,
+            "error": "未指定模块"
+        }, to=sid)
+        return
+
+    try:
+        from src.config.config import cfg
+        diag = await get_diagnostics()
+
+        config_map = {
+            'asr': cfg.asr,
+            'tts': cfg.tts,
+            'ai': cfg.ai,
+            'auto_agent': cfg.auto_agent,
+            'llm_core': cfg.llm_core
+        }
+
+        if module not in config_map:
+            await sio.emit('module_health_result', {
+                "success": False,
+                "error": f"未知模块: {module}"
+            }, to=sid)
+            return
+
+        # 转换配置
+        config_dict = {}
+        if hasattr(config_map[module], '__dataclass_fields__'):
+            from dataclasses import asdict
+            config_dict = asdict(config_map[module])
+        else:
+            config_dict = dict(config_map[module])
+
+        result = await diag.quick_check_module(module, config_dict)
+
+        await sio.emit('module_health_result', {
+            "success": True,
+            "module": module,
+            "data": result.to_dict()
+        }, to=sid)
+
+        # 同时广播给所有客户端更新模块状态
+        await sio.emit('module_status_update', {
+            "module": module,
+            "status": result.to_dict()
+        })
+
+    except Exception as e:
+        await sio.emit('module_health_result', {
+            "success": False,
+            "error": str(e)
+        }, to=sid)
+
+# REST API
+@fastapi_app.get("/api/system/info")
+async def get_system_info():
+    """HTTP备用接口 - 减少日志"""
+    try:
+        cpu = psutil.cpu_percent(interval=0.1)
+        mem = psutil.virtual_memory()
+        disk = psutil.disk_usage('/')
+        proc = psutil.Process()
+
+        return {
+            "success": True,
+            "data": {
+                "cpu": {"percent": cpu, "count": psutil.cpu_count()},
+                "memory": {
+                    "total": mem.total, "available": mem.available,
+                    "percent": mem.percent, "used": mem.used, "free": mem.free
+                },
+                "disk": {
+                    "total": disk.total, "used": disk.used,
+                    "free": disk.free, "percent": (disk.used/disk.total)*100
+                },
+                "process": {
+                    "memory_percent": proc.memory_percent(),
+                    "cpu_percent": proc.cpu_percent(interval=0.1),
+                    "threads": proc.num_threads(),
+                    "uptime": time.time() - proc.create_time()
+                }
+            }
+        }
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+@fastapi_app.get("/api/core/status")
+async def get_core_status_api():
+    """HTTP备用接口"""
+    return {"success": True, "data": get_core_status()}
+
+@fastapi_app.post("/api/core/start")
+async def start_core_api():
+    """HTTP备用接口"""
+    try:
+        status = get_status()
+        if status.is_running:
+            return {"success": True, "data": status.to_dict(), "message": "核心已在运行"}
+
+        success, error = start_core(project_root)
+        if not success:
+            raise HTTPException(status_code=500, detail=error or "启动失败")
+
+        # 等待初始化完成
+        for _ in range(20):
+            await asyncio.sleep(0.5)
+            status = get_status()
+            if status.is_running:
+                # 通过WebSocket广播状态更新（给所有连接的客户端）
+                await sio.emit('core_status', {
+                    'success': True,
+                    'data': status.to_dict(),
+                    'message': '核心启动成功'
+                })
+                return {"success": True, "data": status.to_dict(), "message": "核心启动成功"}
+
+        status = get_status()
+        return {"success": True, "data": status.to_dict(), "message": "核心启动中..."}
+
+    except HTTPException:
+        raise
+    except Exception as e:
+        status = get_status()
+        if status.is_running:
+            return {"success": True, "data": status.to_dict(), "message": "核心已启动"}
+        raise HTTPException(status_code=500, detail=str(e))
+
+@fastapi_app.post("/api/core/stop")
+async def stop_core_api():
+    """HTTP备用接口"""
+    try:
+        status = get_status()
+        if not status.is_running:
+            return {"success": True, "data": status.to_dict(), "message": "核心已停止"}
+
+        success, error = stop_core()
+        if success:
+            for _ in range(20):
+                await asyncio.sleep(0.5)
+                status = get_status()
+                if not status.is_running:
+                    # 广播状态更新
+                    await sio.emit('core_status', {
+                        'success': True,
+                        'data': status.to_dict(),
+                        'message': '核心已停止'
+                    })
+                    return {"success": True, "data": status.to_dict(), "message": "核心已停止"}
+
+            status = get_status()
+            return {"success": True, "data": status.to_dict(), "message": error or "核心停止中..."}
+        else:
+            raise HTTPException(status_code=500, detail=error or "停止失败")
+
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+@fastapi_app.get("/api/modules/status")
+async def get_modules_status():
+    """模块状态"""
+    try:
+        from src.config.config import cfg
+        return {
+            "success": True,
+            "data": {
+                "asr": {"enabled": cfg.asr.enabled},
+                "tts": {"enabled": cfg.tts.enabled},
+                "ai": {"enabled": cfg.ai.api_key is not None},
+                "auto_agent": {"enabled": cfg.auto_agent.enabled},
+                "llm_core": {"enabled": cfg.llm_core.enabled}
+            }
+        }
+    except:
+        return {
+            "success": True,
+            "data": {
+                "asr": {"enabled": True}, "tts": {"enabled": True},
+                "ai": {"enabled": True}, "auto_agent": {"enabled": True},
+                "llm_core": {"enabled": True}
+            }
+        }
+
+@fastapi_app.get("/api/config")
+async def get_config():
+    from src.config.config import cfg
+    return {"success": True, "data": cfg.to_dict()}
+
+@fastapi_app.post("/api/config/{section}")
+async def update_config(section: str, data: Dict[str, Any]):
+    from src.config.config import cfg
+    cfg.update({section: data})
+    return {"success": True, "message": "配置已更新"}
+
+@fastapi_app.post("/api/config/reload")
+async def reload_config():
+    from src.config.config import cfg
+    cfg.reload()
+    return {"success": True, "message": "配置已重载"}
+
+@fastapi_app.get("/api/preferences")
+async def get_preferences():
+    prefs_path = Path(__file__).parent / "user_preferences.json"
+    if prefs_path.exists():
+        with open(prefs_path, 'r', encoding='utf-8') as f:
+            return {"success": True, "data": json.load(f)}
+    return {"success": True, "data": {}}
+
+@fastapi_app.post("/api/preferences")
+async def save_preferences(data: Dict[str, Any]):
+    prefs_path = Path(__file__).parent / "user_preferences.json"
+    with open(prefs_path, 'w', encoding='utf-8') as f:
+        json.dump(data, f, indent=2, ensure_ascii=False)
+    return {"success": True, "message": "偏好已保存"}
+
+# 静态文件
+static_dir = Path(__file__).parent / "static"
+if static_dir.exists():
+    fastapi_app.mount("/", StaticFiles(directory=static_dir, html=True), name="static")
+
+def run_server(host: str = "0.0.0.0", port: int = 8089, debug: bool = False):
+    import uvicorn
+    uvicorn.run("backend.app:app", host=host, port=port, reload=debug,
+                access_log=False)  # 禁用默认访问日志（我们使用自定义过滤器）
+
+if __name__ == "__main__":
+    run_server()

src/server_view/backend/core_manager.py

@@ -0,0 +1,248 @@
+"""
+Yosuga Server 核心进程管理器
+"""
+import asyncio
+import threading
+import time
+import sys
+import logging
+from pathlib import Path
+from typing import Optional, Dict, Any
+from dataclasses import dataclass
+
+# 全局状态
+_core_instance: Optional[Any] = None
+_core_thread: Optional[threading.Thread] = None
+_core_start_time: Optional[float] = None
+_core_stop_event: threading.Event = threading.Event()  # 停止信号
+_core_lock = threading.Lock()
+_logger: Optional[logging.Logger] = None
+
+
+@dataclass
+class CoreStatus:
+    is_running: bool = False
+    pid: int = 0
+    uptime: float = 0.0
+    error: Optional[str] = None
+    thread_alive: bool = False
+    start_time: Optional[str] = None
+
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "is_running": self.is_running,
+            "pid": self.pid,
+            "uptime": round(self.uptime, 1),
+            "error": self.error,
+            "thread_alive": self.thread_alive,
+            "start_time": self.start_time
+        }
+
+import os
+def get_status() -> CoreStatus:
+    """获取当前核心状态"""
+    global _core_thread, _core_start_time, _core_instance
+
+    with _core_lock:
+        status = CoreStatus()
+
+        if _core_thread is not None:
+            status.thread_alive = _core_thread.is_alive()
+            status.pid = os.getpid()
+
+        if _core_start_time and (status.thread_alive or _core_instance is not None):
+            status.uptime = time.time() - _core_start_time
+            status.is_running = True
+            status.start_time = time.strftime("%Y-%m-%dT%H:%M:%S", time.localtime(_core_start_time))
+
+        return status
+
+
+def _setup_loguru_logging(log_queue: Optional[Any] = None):
+    """配置loguru，同时转发到标准logging以便Socket.IO捕获"""
+    try:
+        from loguru import logger
+        import logging
+
+        logger.remove()
+
+        # 关键：添加一个sink将loguru日志转发到标准logging
+        class LoguruToStandard:
+            def write(self, message):
+                # 解析loguru格式提取level
+                record = message.record
+                level = record["level"].name
+                # 获取标准logger并发送
+                std_logger = logging.getLogger("yosuga")
+                if level == "DEBUG":
+                    std_logger.debug(record["message"])
+                elif level == "INFO":
+                    std_logger.info(record["message"])
+                elif level == "SUCCESS":
+                    std_logger.info(record["message"])
+                elif level == "WARNING":
+                    std_logger.warning(record["message"])
+                elif level == "ERROR":
+                    std_logger.error(record["message"])
+                elif level == "CRITICAL":
+                    std_logger.critical(record["message"])
+
+            def flush(self):
+                pass
+
+        # 添加转发处理器
+        logger.add(LoguruToStandard(), format="{message}")
+
+        # 文件日志
+        from src.config.config import cfg
+        log_dir = Path(cfg.log_dir)
+        log_dir.mkdir(parents=True, exist_ok=True)
+        logger.add(
+            f"{log_dir}/Yosuga_server-{{time:YYYY-MM-DD_HH-mm-ss}}.log",
+            encoding="utf-8",
+            rotation="100 MB"
+        )
+
+        return logger
+    except Exception as e:
+        print(f"Loguru配置错误: {e}")
+        return None
+
+
+def _run_core_thread(project_root: Path):
+    """在独立线程中运行YosugaServerCore"""
+    global _core_instance, _core_start_time, _core_stop_event, _logger
+
+    _core_stop_event.clear()
+
+    try:
+        sys.path.insert(0, str(project_root))
+
+        from src.server_core.core import YosugaServerCore
+        from src.config.config import cfg
+
+        # 配置loguru并获取logger
+        logger = _setup_loguru_logging()
+        if logger:
+            logger.info("Yosuga_server 在线程中启动")
+            _logger = logger
+
+        async def run_core():
+            global _core_instance
+            try:
+                _core_instance = await YosugaServerCore.get_instance()
+
+                if logger:
+                    logger.success(f"YosugaServerCore 初始化完成，线程ID: {threading.current_thread().ident}")
+
+                # 运行核心，同时检查停止信号
+                core_task = asyncio.create_task(_core_instance.run())
+
+                # 等待任务完成或收到停止信号
+                while not core_task.done():
+                    if _core_stop_event.is_set():
+                        core_task.cancel()
+                        try:
+                            await core_task
+                        except asyncio.CancelledError:
+                            if logger:
+                                logger.info("核心收到停止信号，正在关闭...")
+                        break
+                    await asyncio.sleep(0.1)
+
+                if logger:
+                    logger.info("核心事件循环已结束")
+
+            except asyncio.CancelledError:
+                if logger:
+                    logger.info("核心任务已取消")
+            except Exception as e:
+                if logger:
+                    logger.exception(f"核心运行异常: {e}")
+                raise
+
+        # 运行异步核心
+        asyncio.run(run_core())
+
+    except Exception as e:
+        import traceback
+        error_msg = f"核心线程异常: {str(e)}\n{traceback.format_exc()}"
+        print(error_msg, file=sys.stderr)
+        if _logger:
+            _logger.error(error_msg)
+    finally:
+        _core_instance = None
+        if logger:
+            logger.info("核心线程已退出")
+
+
+def start_core(project_root: Path) -> tuple[bool, Optional[str]]:
+    """启动Yosuga核心"""
+    global _core_thread, _core_start_time, _core_stop_event
+
+    with _core_lock:
+        if _core_thread is not None and _core_thread.is_alive():
+            return True, None  # 已经在运行
+
+        _core_start_time = None
+        _core_stop_event.clear()
+
+        try:
+            _core_thread = threading.Thread(
+                target=_run_core_thread,
+                args=(project_root,),
+                name="YosugaServerCore",
+                daemon=True
+            )
+            _core_thread.start()
+            _core_start_time = time.time()  # 立即记录
+
+            time.sleep(0.5)
+
+            if not _core_thread.is_alive():
+                _core_start_time = None
+                return False, "核心线程未能启动"
+
+            return True, None
+
+        except Exception as e:
+            _core_start_time = None
+            return False, str(e)
+
+
+def stop_core() -> tuple[bool, Optional[str]]:
+    """停止Yosuga核心"""
+    global _core_thread, _core_start_time, _core_stop_event
+
+    with _core_lock:
+        if _core_thread is None or not _core_thread.is_alive():
+            _core_thread = None
+            _core_start_time = None
+            return True, None  # 已经停止
+
+        try:
+            # 发送停止信号
+            _core_stop_event.set()
+
+            # 等待线程结束（带超时）
+            _core_thread.join(timeout=10.0)
+
+            was_alive = _core_thread.is_alive()
+            _core_thread = None
+
+            if was_alive:
+                # 线程还在运行，但已经发送了停止信号
+                # 由于daemon=True，主进程退出时会强制终止
+                _core_start_time = None
+                return True, "核心停止信号已发送，正在后台停止"
+
+            _core_start_time = None
+            return True, None
+
+        except Exception as e:
+            return False, str(e)
+
+
+# 兼容旧接口
+def get_core_status() -> Dict[str, Any]:
+    return get_status().to_dict()

src/server_view/backend/diagnostics.py

@@ -0,0 +1,524 @@
+"""
+Yosuga Server 系统诊断模块 - TCP端口连通性版本
+生产级健康检查与自检工具（不依赖HTTP接口）
+"""
+import asyncio
+import json
+import socket
+import time
+from dataclasses import dataclass, field
+from enum import Enum
+from pathlib import Path
+from typing import Dict, List, Optional, Any, Tuple
+from urllib.parse import urlparse
+import psutil
+from loguru import logger
+
+class HealthStatus(Enum):
+    HEALTHY = "healthy"
+    UNHEALTHY = "unhealthy"
+    UNKNOWN = "unknown"
+    CHECKING = "checking"
+
+@dataclass
+class CheckResult:
+    """单项检查结果"""
+    name: str
+    status: HealthStatus
+    message: str
+    details: Dict[str, Any] = field(default_factory=dict)
+    latency_ms: float = 0.0
+    timestamp: float = field(default_factory=time.time)
+
+    def to_dict(self) -> dict:
+        return {
+            "name": self.name,
+            "status": self.status.value,
+            "message": self.message,
+            "details": self.details,
+            "latency_ms": round(self.latency_ms, 2),
+            "timestamp": self.timestamp
+        }
+
+@dataclass
+class DiagnosticsReport:
+    """完整诊断报告"""
+    overall_status: HealthStatus
+    checks: List[CheckResult]
+    summary: Dict[str, int]
+    generated_at: float = field(default_factory=time.time)
+    version: str = "1.1.0"
+
+    def to_dict(self) -> dict:
+        return {
+            "overall_status": self.overall_status.value,
+            "checks": [c.to_dict() for c in self.checks],
+            "summary": self.summary,
+            "generated_at": self.generated_at,
+            "version": self.version
+        }
+
+class SystemDiagnostics:
+    """系统诊断核心类 - TCP端口连通性检测"""
+
+    def __init__(self, config_path: Optional[Path] = None):
+        self.config_path = config_path or self._find_config_path()
+        self._timeout_seconds = 3
+
+    def _find_config_path(self) -> Path:
+        """自动查找配置文件路径"""
+        markers = ['settings.json', 'pyproject.toml']
+        current = Path(__file__).resolve().parent.parent.parent.parent
+
+        for path in [current, *current.parents]:
+            if (path / 'settings.json').exists():
+                return path / 'settings.json'
+            if path == path.parent:
+                break
+        return current / 'settings.json'
+
+    async def _check_tcp_port(self, host: str, port: int) -> Tuple[bool, float, Optional[str]]:
+        """
+        基础TCP端口连通性检查
+
+        Returns:
+            (是否连通, 延迟ms, 错误信息)
+        """
+        start = time.time()
+        try:
+            reader, writer = await asyncio.wait_for(
+                asyncio.open_connection(host, port),
+                timeout=self._timeout_seconds
+            )
+            writer.close()
+            await writer.wait_closed()
+            latency = (time.time() - start) * 1000
+            return True, latency, None
+        except asyncio.TimeoutError:
+            return False, self._timeout_seconds * 1000, "连接超时"
+        except ConnectionRefusedError:
+            return False, (time.time() - start) * 1000, "连接被拒绝"
+        except Exception as e:
+            return False, (time.time() - start) * 1000, str(e)
+
+    def _parse_url(self, url: str) -> Tuple[str, int]:
+        """
+        从URL解析主机和端口
+        支持 http://host:port/path 格式
+        """
+        try:
+            parsed = urlparse(url)
+            host = parsed.hostname or 'localhost'
+            if parsed.port:
+                port = parsed.port
+            elif parsed.scheme == 'https':
+                port = 443
+            elif parsed.scheme == 'http':
+                port = 80
+            else:
+                port = 80
+            return host, port
+        except Exception:
+            if ':' in url:
+                parts = url.split(':')
+                if len(parts) >= 2:
+                    last_part = parts[-1].split('/')[0]
+                    try:
+                        return parts[-2].replace('//', ''), int(last_part)
+                    except:
+                        pass
+            return 'localhost', 80
+
+    async def check_asr(self, url: str = "http://localhost:20260") -> CheckResult:
+        """检查ASR服务 - TCP端口连通性"""
+        host, port = self._parse_url(url)
+        if port == 80 and '20260' in url:
+            port = 20260
+
+        is_open, latency, error = await self._check_tcp_port(host, port)
+
+        if is_open:
+            return CheckResult(
+                name="ASR服务",
+                status=HealthStatus.HEALTHY,
+                message=f"端口可连通 {host}:{port}",
+                details={"host": host, "port": port, "protocol": "TCP"},
+                latency_ms=latency
+            )
+        else:
+            return CheckResult(
+                name="ASR服务",
+                status=HealthStatus.UNHEALTHY,
+                message=f"端口不可达 {host}:{port} - {error}",
+                details={"host": host, "port": port, "error": error},
+                latency_ms=latency
+            )
+
+    async def check_tts(self, host: str = "localhost", port: int = 20261) -> CheckResult:
+        """检查TTS服务 - TCP端口连通性"""
+        is_open, latency, error = await self._check_tcp_port(host, port)
+
+        if is_open:
+            return CheckResult(
+                name="TTS服务",
+                status=HealthStatus.HEALTHY,
+                message=f"端口可连通 {host}:{port}",
+                details={"host": host, "port": port},
+                latency_ms=latency
+            )
+        else:
+            return CheckResult(
+                name="TTS服务",
+                status=HealthStatus.UNHEALTHY,
+                message=f"端口不可达 {host}:{port} - {error}",
+                details={"host": host, "port": port, "error": error},
+                latency_ms=latency
+            )
+
+    async def check_ai_service(self, base_url: str, api_key: Optional[str] = None) -> CheckResult:
+        """检查AI服务 - TCP端口连通性"""
+        host, port = self._parse_url(base_url)
+
+        is_open, latency, error = await self._check_tcp_port(host, port)
+
+        if is_open:
+            return CheckResult(
+                name="AI服务",
+                status=HealthStatus.HEALTHY,
+                message=f"端口可连通 {host}:{port}",
+                details={"host": host, "port": port, "base_url": base_url},
+                latency_ms=latency
+            )
+        else:
+            return CheckResult(
+                name="AI服务",
+                status=HealthStatus.UNHEALTHY,
+                message=f"端口不可达 {host}:{port} - {error}",
+                details={"host": host, "port": port, "error": error},
+                latency_ms=latency
+            )
+
+    async def check_auto_agent(self, base_url: str) -> CheckResult:
+        """检查Auto Agent服务 - TCP端口连通性"""
+        host, port = self._parse_url(base_url)
+
+        is_open, latency, error = await self._check_tcp_port(host, port)
+
+        if is_open:
+            return CheckResult(
+                name="自动代理服务",
+                status=HealthStatus.HEALTHY,
+                message=f"端口可连通 {host}:{port}",
+                details={"host": host, "port": port},
+                latency_ms=latency
+            )
+        else:
+            return CheckResult(
+                name="自动代理服务",
+                status=HealthStatus.UNHEALTHY,
+                message=f"端口不可达 {host}:{port} - {error}",
+                details={"host": host, "port": port, "error": error},
+                latency_ms=latency
+            )
+
+    async def check_config_file(self) -> CheckResult:
+        """检查配置文件合法性"""
+        try:
+            if not self.config_path.exists():
+                return CheckResult(
+                    name="配置文件",
+                    status=HealthStatus.UNHEALTHY,
+                    message=f"配置文件不存在: {self.config_path}"
+                )
+
+            with open(self.config_path, 'r', encoding='utf-8') as f:
+                content = f.read()
+                config = json.loads(content)
+
+            required_sections = ['ai', 'tts', 'asr']
+            missing = [s for s in required_sections if s not in config]
+
+            if missing:
+                return CheckResult(
+                    name="配置文件",
+                    status=HealthStatus.UNHEALTHY,
+                    message=f"缺少配置节: {', '.join(missing)}",
+                    details={"missing_sections": missing}
+                )
+
+            return CheckResult(
+                name="配置文件",
+                status=HealthStatus.HEALTHY,
+                message=f"配置合法，含 {len(config)} 个配置节",
+                details={"sections": list(config.keys())}
+            )
+
+        except json.JSONDecodeError as e:
+            return CheckResult(
+                name="配置文件",
+                status=HealthStatus.UNHEALTHY,
+                message=f"JSON格式错误: {str(e)}"
+            )
+        except Exception as e:
+            return CheckResult(
+                name="配置文件",
+                status=HealthStatus.UNHEALTHY,
+                message=f"读取失败: {str(e)}"
+            )
+
+    async def check_model_files(self, config: Optional[Dict] = None) -> CheckResult:
+        """检查模型文件存在性"""
+        try:
+            if config is None:
+                with open(self.config_path, 'r', encoding='utf-8') as f:
+                    config = json.load(f)
+
+            checks = {}
+            missing = []
+            project_root = self.config_path.parent
+
+            if 'tts' in config:
+                tts = config['tts']
+                gpt_path = tts.get('gpt_model_name', '')
+                sovits_path = tts.get('sovits_model_name', '')
+
+                if gpt_path:
+                    full_path = project_root / gpt_path
+                    exists = full_path.exists()
+                    checks['gpt_model'] = {"path": str(full_path), "exists": exists}
+                    if not exists:
+                        missing.append(f"GPT模型: {gpt_path}")
+
+                if sovits_path:
+                    full_path = project_root / sovits_path
+                    exists = full_path.exists()
+                    checks['sovits_model'] = {"path": str(full_path), "exists": exists}
+                    if not exists:
+                        missing.append(f"SoVITS模型: {sovits_path}")
+
+            if missing:
+                return CheckResult(
+                    name="模型文件",
+                    status=HealthStatus.UNHEALTHY,
+                    message=f"缺少 {len(missing)} 个模型文件",
+                    details={"missing": missing, "checks": checks}
+                )
+
+            return CheckResult(
+                name="模型文件",
+                status=HealthStatus.HEALTHY,
+                message="所有配置模型文件已找到",
+                details={"checks": checks}
+            )
+
+        except Exception as e:
+            return CheckResult(
+                name="模型文件",
+                status=HealthStatus.UNKNOWN,
+                message=f"检查失败: {str(e)}"
+            )
+
+    async def check_ports(self, ports: Optional[List[int]] = None) -> CheckResult:
+        """检查关键端口占用情况"""
+        if ports is None:
+            ports = [8089, 20260, 20261, 8765]
+
+        try:
+            current_pid = psutil.Process().pid
+            current_ports = set()
+
+            proc = psutil.Process(current_pid)
+            for conn in proc.connections(kind='inet'):
+                if conn.status == 'LISTEN':
+                    current_ports.add(conn.lport)
+
+            occupied_by_others = []
+            for port in ports:
+                if port in current_ports:
+                    continue
+
+                for p in psutil.process_iter(['pid', 'name']):
+                    try:
+                        for conn in p.connections(kind='inet'):
+                            if conn.lport == port:
+                                occupied_by_others.append({
+                                    "port": port,
+                                    "pid": p.pid,
+                                    "name": p.name()
+                                })
+                                break
+                    except (psutil.NoSuchProcess, psutil.AccessDenied):
+                        continue
+
+            if occupied_by_others:
+                return CheckResult(
+                    name="端口占用",
+                    status=HealthStatus.UNHEALTHY,
+                    message=f"{len(occupied_by_others)} 个端口被占用",
+                    details={"conflicts": occupied_by_others}
+                )
+
+            return CheckResult(
+                name="端口占用",
+                status=HealthStatus.HEALTHY,
+                message="关键端口可用",
+                details={"checked_ports": list(ports), "self_ports": list(current_ports)}
+            )
+
+        except Exception as e:
+            return CheckResult(
+                name="端口占用",
+                status=HealthStatus.UNKNOWN,
+                message=f"检查失败: {str(e)}"
+            )
+
+    async def run_full_diagnostics(self) -> DiagnosticsReport:
+        """执行完整系统体检"""
+        logger.info("开始系统体检(TCP模式)...")
+        checks = []
+
+        checks.append(await self.check_config_file())
+        checks.append(await self.check_ports())
+
+        service_results = await self._check_services_tcp()
+        checks.extend(service_results)
+
+        checks.append(await self.check_model_files())
+
+        summary = {
+            "healthy": sum(1 for c in checks if c.status == HealthStatus.HEALTHY),
+            "unhealthy": sum(1 for c in checks if c.status == HealthStatus.UNHEALTHY),
+            "unknown": sum(1 for c in checks if c.status == HealthStatus.UNKNOWN),
+            "total": len(checks)
+        }
+
+        if summary["unhealthy"] == 0:
+            overall = HealthStatus.HEALTHY
+        elif summary["unhealthy"] <= summary["healthy"]:
+            overall = HealthStatus.UNKNOWN
+        else:
+            overall = HealthStatus.UNHEALTHY
+
+        report = DiagnosticsReport(
+            overall_status=overall,
+            checks=checks,
+            summary=summary
+        )
+
+        logger.info(f"体检完成: {summary['healthy']}/{summary['total']} 项正常")
+        return report
+
+    async def _check_services_tcp(self) -> List[CheckResult]:
+        """检查各项服务 - 纯TCP连通性"""
+        results = []
+
+        try:
+            with open(self.config_path, 'r', encoding='utf-8') as f:
+                config = json.load(f)
+
+            if 'asr' in config:
+                asr_url = config['asr'].get('url', 'http://localhost:20260')
+                results.append(await self.check_asr(asr_url))
+            else:
+                results.append(CheckResult(
+                    name="ASR服务",
+                    status=HealthStatus.UNKNOWN,
+                    message="配置中未启用ASR"
+                ))
+
+            if 'tts' in config:
+                tts_cfg = config['tts']
+                results.append(await self.check_tts(
+                    tts_cfg.get('host', 'localhost'),
+                    tts_cfg.get('port', 20261)
+                ))
+            else:
+                results.append(CheckResult(
+                    name="TTS服务",
+                    status=HealthStatus.UNKNOWN,
+                    message="配置中未启用TTS"
+                ))
+
+            if 'ai' in config:
+                ai_cfg = config['ai']
+                results.append(await self.check_ai_service(
+                    ai_cfg.get('base_url', 'http://localhost:1234/v1')
+                ))
+            else:
+                results.append(CheckResult(
+                    name="AI服务",
+                    status=HealthStatus.UNKNOWN,
+                    message="配置中未启用AI"
+                ))
+
+            if 'auto_agent' in config:
+                aa_cfg = config['auto_agent']
+                results.append(await self.check_auto_agent(
+                    aa_cfg.get('base_url', 'http://localhost:1234/v1')
+                ))
+            else:
+                results.append(CheckResult(
+                    name="自动代理服务",
+                    status=HealthStatus.UNKNOWN,
+                    message="配置中未启用自动代理"
+                ))
+
+        except Exception as e:
+            logger.error(f"服务检查失败: {e}")
+            results.append(CheckResult(
+                name="服务检查",
+                status=HealthStatus.UNHEALTHY,
+                message=f"配置加载失败: {str(e)}"
+            ))
+
+        return results
+
+    async def quick_check_module(self, module_name: str, config: Dict) -> CheckResult:
+        """快速检查单个模块 - TCP连通性"""
+        if module_name == "asr":
+            url = config.get('url', 'http://localhost:20260')
+            return await self.check_asr(url)
+        elif module_name == "tts":
+            return await self.check_tts(
+                config.get('host', 'localhost'),
+                config.get('port', 20261)
+            )
+        elif module_name == "ai":
+            return await self.check_ai_service(
+                config.get('base_url', 'http://localhost:1234/v1')
+            )
+        elif module_name == "auto_agent":
+            return await self.check_auto_agent(
+                config.get('base_url', 'http://localhost:1234/v1')
+            )
+        elif module_name == "llm_core":
+            from src.server_view.backend.core_manager import get_status
+            status = get_status()
+            if status.is_running:
+                return CheckResult(
+                    name="LLM核心",
+                    status=HealthStatus.HEALTHY,
+                    message="核心进程运行中",
+                    details={"uptime": status.uptime, "pid": status.pid}
+                )
+            else:
+                return CheckResult(
+                    name="LLM核心",
+                    status=HealthStatus.UNHEALTHY,
+                    message="核心进程未启动"
+                )
+        else:
+            return CheckResult(
+                name=module_name,
+                status=HealthStatus.UNKNOWN,
+                message="未知模块"
+            )
+
+_diagnostics_instance: Optional[SystemDiagnostics] = None
+
+async def get_diagnostics() -> SystemDiagnostics:
+    """获取诊断实例（单例）"""
+    global _diagnostics_instance
+    if _diagnostics_instance is None:
+        _diagnostics_instance = SystemDiagnostics()
+    return _diagnostics_instance

src/server_view/frontend/.gitignore

@@ -0,0 +1,21 @@
+# Python-generated files
+__pycache__/
+*.py[oc]
+build/
+dist/
+wheels/
+*.egg-info
+
+# from ide
+.vscode
+.idea
+
+# Virtual environments
+.venv
+venv
+
+node_modules
+node_modules/
+dist
+dist/
+pnpm-lock.yaml

src/server_view/frontend/index.html

@@ -0,0 +1,13 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+  <head>
+    <meta charset="UTF-8" />
+    <link rel="icon" type="image/svg+xml" href="/vite.svg" />
+    <meta name="viewport" content="width=device-width, initial-scale=1.0" />
+    <title>Yosuga Server - 管理界面</title>
+  </head>
+  <body>
+    <div id="app"></div>
+    <script type="module" src="/src/main.ts"></script>
+  </body>
+</html>

src/server_view/frontend/package.json

@@ -0,0 +1,29 @@
+{
+  "name": "yosuga-server-web",
+  "private": true,
+  "version": "0.1.0",
+  "type": "module",
+  "scripts": {
+    "dev": "vite",
+    "build": "vue-tsc && vite build",
+    "preview": "vite preview"
+  },
+  "dependencies": {
+    "@vicons/ionicons5": "^0.12.0",
+    "axios": "^1.6.0",
+    "echarts": "^5.4.3",
+    "naive-ui": "^2.35.0",
+    "pinia": "^2.1.7",
+    "socket.io-client": "^4.7.2",
+    "vue": "^3.3.8",
+    "vue-echarts": "^6.6.1",
+    "vue-router": "^4.2.5"
+  },
+  "devDependencies": {
+    "@types/node": "^20.9.0",
+    "@vitejs/plugin-vue": "^4.5.0",
+    "typescript": "^5.2.2",
+    "vite": "^5.0.0",
+    "vue-tsc": "^1.8.22"
+  }
+}

src/server_view/frontend/src/App.vue

@@ -0,0 +1,238 @@
+<template>
+  <n-config-provider :theme="currentTheme" :theme-overrides="themeOverrides">
+    <n-loading-bar-provider>
+      <n-dialog-provider>
+        <n-notification-provider>
+          <n-message-provider>
+            <n-layout class="layout" position="absolute" has-sider>
+              <!-- 侧边栏 -->
+              <n-layout-sider
+                bordered
+                collapse-mode="width"
+                :collapsed-width="64"
+                :width="240"
+                :collapsed="prefs.preferences.sidebarCollapsed"
+                show-trigger
+                @collapse="prefs.toggleSidebar()"
+                @expand="prefs.toggleSidebar()"
+              >
+                <div class="logo">
+                  <n-icon size="32" color="#18a058">
+                    <logo-icon />
+                  </n-icon>
+                  <span v-if="!prefs.preferences.sidebarCollapsed" class="logo-text">Yosuga Server</span>
+                </div>
+
+                <n-menu
+                  :collapsed="prefs.preferences.sidebarCollapsed"
+                  :collapsed-width="64"
+                  :collapsed-icon-size="22"
+                  :options="menuOptions"
+                  :value="activeKey"
+                  @update:value="handleMenuUpdate"
+                />
+              </n-layout-sider>
+
+              <!-- 主内容区 -->
+              <n-layout>
+                <n-layout-header bordered class="header">
+                  <div class="header-left">
+                    <n-breadcrumb>
+                      <n-breadcrumb-item>Yosuga</n-breadcrumb-item>
+                      <n-breadcrumb-item>{{ pageTitle }}</n-breadcrumb-item>
+                    </n-breadcrumb>
+                  </div>
+                  <div class="header-right">
+                    <!-- 主题切换 -->
+                    <n-dropdown :options="themeOptions" @select="handleThemeSelect">
+                      <n-button circle>
+                        <template #icon>
+                          <n-icon>
+                            <sunny-icon v-if="prefs.preferences.theme === 'light'" />
+                            <moon-icon v-else-if="prefs.preferences.theme === 'dark'" />
+                            <contrast-icon v-else />
+                          </n-icon>
+                        </template>
+                      </n-button>
+                    </n-dropdown>
+
+                    <n-badge :value="wsStatus === 'connected' ? 'Live' : 'Offline'"
+                             :type="wsStatus === 'connected' ? 'success' : 'error'">
+                      <n-button circle>
+                        <template #icon>
+                          <n-icon><wifi-icon /></n-icon>
+                        </template>
+                      </n-button>
+                    </n-badge>
+                  </div>
+                </n-layout-header>
+
+                <n-layout-content class="content">
+                  <router-view v-slot="{ Component }">
+                    <transition name="fade" mode="out-in">
+                      <component :is="Component" />
+                    </transition>
+                  </router-view>
+                </n-layout-content>
+              </n-layout>
+            </n-layout>
+          </n-message-provider>
+        </n-notification-provider>
+      </n-dialog-provider>
+    </n-loading-bar-provider>
+  </n-config-provider>
+</template>
+
+<script setup lang="ts">
+import { ref, computed, h, onMounted } from 'vue'
+import { useRoute, useRouter } from 'vue-router'
+import { darkTheme, lightTheme, type GlobalTheme, type DropdownOption } from 'naive-ui'
+import {
+  HomeOutline as HomeIcon,
+  SettingsOutline as SettingsIcon,
+  TerminalOutline as TerminalIcon,
+  StatsChartOutline as StatsIcon,
+  HardwareChipOutline as DeviceIcon,
+  SunnyOutline as SunnyIcon,
+  MoonOutline as MoonIcon,
+  WifiOutline as WifiIcon,
+  PulseOutline as LogoIcon,
+  ContrastOutline as ContrastIcon
+} from '@vicons/ionicons5'
+import { useWebSocketStore } from '@/stores/websocket'
+import { usePreferencesStore } from '@/stores/preferences'
+import { storeToRefs } from 'pinia'
+import { NIcon } from 'naive-ui'
+
+const route = useRoute()
+const router = useRouter()
+const wsStore = useWebSocketStore()
+const { status: wsStatus } = storeToRefs(wsStore)
+const prefs = usePreferencesStore()
+
+const activeKey = computed(() => route.name as string)
+const pageTitle = computed(() => route.meta?.title as string || 'Dashboard')
+
+// 主题计算
+const currentTheme = computed<GlobalTheme | null>(() => {
+  if (prefs.preferences.theme === 'dark') return darkTheme
+  if (prefs.preferences.theme === 'light') return lightTheme
+  return null // auto 时由 CSS 处理
+})
+
+const themeOverrides = {
+  common: {
+    primaryColor: '#18a058',
+    primaryColorHover: '#36ad6a',
+    primaryColorPressed: '#0c7a43',
+  }
+}
+
+// 主题选项
+const themeOptions: DropdownOption[] = [
+  { label: '浅色模式', key: 'light', icon: () => h(NIcon, null, { default: () => h(SunnyIcon) }) },
+  { label: '深色模式', key: 'dark', icon: () => h(NIcon, null, { default: () => h(MoonIcon) }) },
+  { label: '跟随系统', key: 'auto', icon: () => h(NIcon, null, { default: () => h(ContrastIcon) }) }
+]
+
+const renderIcon = (icon: any) => {
+  return () => h(NIcon, null, { default: () => h(icon) })
+}
+
+const menuOptions = [
+  { label: '仪表板', key: 'dashboard', icon: renderIcon(HomeIcon) },
+  { label: '系统监控', key: 'monitor', icon: renderIcon(StatsIcon) },
+  { label: '设备管理', key: 'devices', icon: renderIcon(DeviceIcon) },
+  { label: '配置管理', key: 'config', icon: renderIcon(SettingsIcon) },
+  { label: '日志查看', key: 'logs', icon: renderIcon(TerminalIcon) }
+]
+
+const handleMenuUpdate = (key: string) => {
+  prefs.preferences.lastVisitedPage = key
+  router.push({ name: key })
+}
+
+const handleThemeSelect = (key: string) => {
+  prefs.setTheme(key)
+}
+
+onMounted(() => {
+  prefs.loadPreferences()
+  wsStore.connect()
+
+  // 恢复上次访问的页面
+  if (prefs.preferences.lastVisitedPage && prefs.preferences.lastVisitedPage !== 'dashboard') {
+    router.push({ name: prefs.preferences.lastVisitedPage })
+  }
+})
+</script>
+
+<style>
+/* 全局主题 CSS 变量 */
+:root {
+  color-scheme: light dark;
+}
+
+/* 深色模式自动适配 */
+@media (prefers-color-scheme: dark) {
+  :root:not(.light) {
+    color-scheme: dark;
+  }
+}
+
+:root.dark {
+  color-scheme: dark;
+}
+</style>
+
+<style scoped>
+.layout {
+  height: 100vh;
+}
+
+.logo {
+  height: 64px;
+  display: flex;
+  align-items: center;
+  justify-content: center;
+  gap: 12px;
+  border-bottom: 1px solid var(--n-border-color);
+}
+
+.logo-text {
+  font-size: 18px;
+  font-weight: 600;
+  color: var(--n-text-color);
+  white-space: nowrap;
+}
+
+.header {
+  height: 64px;
+  padding: 0 24px;
+  display: flex;
+  align-items: center;
+  justify-content: space-between;
+}
+
+.header-right {
+  display: flex;
+  gap: 12px;
+  align-items: center;
+}
+
+.content {
+  padding: 24px;
+  background-color: var(--n-color);
+  overflow-y: auto;
+}
+
+.fade-enter-active,
+.fade-leave-active {
+  transition: opacity 0.3s ease;
+}
+
+.fade-enter-from,
+.fade-leave-to {
+  opacity: 0;
+}
+</style>

src/server_view/frontend/src/components/CoreControl.vue

@@ -0,0 +1,278 @@
+<template>
+  <n-card class="core-control" :bordered="false">
+    <n-space align="center" justify="space-between">
+      <n-space align="center">
+        <n-badge
+          dot
+          :type="isRunning ? 'success' : 'error'"
+          :processing="isLoading"
+        >
+          <n-icon size="28" :color="getStatusColor">
+            <pulse-icon />
+          </n-icon>
+        </n-badge>
+        <div>
+          <div class="core-title">
+            Yosuga 核心
+            <n-tag v-if="isRunning" type="success" size="small" style="margin-left: 8px">
+              运行中
+            </n-tag>
+            <n-tag v-else-if="hasError" type="error" size="small" style="margin-left: 8px">
+              启动失败
+            </n-tag>
+            <n-tag v-else type="default" size="small" style="margin-left: 8px">
+              已停止
+            </n-tag>
+          </div>
+          <div class="core-status">
+            <span v-if="isRunning">
+              PID: {{ coreStatus?.pid || 0 }} |
+              运行时间: {{ formatUptime(coreStatus?.uptime || 0) }} |
+              线程{{ coreStatus?.thread_alive ? '存活' : '异常' }}
+            </span>
+            <span v-else-if="hasError" class="error-text">
+              错误: {{ coreStatus?.error }}
+            </span>
+            <span v-else>点击启动按钮开始运行</span>
+          </div>
+        </div>
+      </n-space>
+
+      <n-space>
+        <!-- 启动按钮 -->
+        <n-button
+          v-if="!isRunning && !isLoading"
+          type="primary"
+          size="large"
+          @click="handleStart"
+        >
+          <template #icon>
+            <n-icon><play-icon /></n-icon>
+          </template>
+          启动核心
+        </n-button>
+
+        <!-- 停止按钮 -->
+        <n-button
+          v-if="isRunning"
+          type="error"
+          size="large"
+          :loading="isLoading"
+          @click="handleStop"
+        >
+          <template #icon>
+            <n-icon><stop-icon /></n-icon>
+          </template>
+          停止核心
+        </n-button>
+
+        <!-- 刷新按钮（现在主要用于强制刷新HTTP备用） -->
+        <n-button circle size="large" :loading="isLoading" @click="refreshStatus">
+          <template #icon>
+            <n-icon><refresh-icon /></n-icon>
+          </template>
+        </n-button>
+      </n-space>
+    </n-space>
+
+    <!-- WebSocket连接状态指示器 -->
+    <n-divider />
+    <n-space align="center" justify="space-between" size="small">
+      <n-space align="center" size="small">
+        <n-badge :type="wsStore.status === 'connected' ? 'success' : 'error'" dot />
+        <span style="font-size: 12px; color: var(--n-text-color-3)">
+          {{ wsStore.status === 'connected' ? '实时连接正常' : '实时连接断开' }}
+          <span v-if="wsStore.latency > 0">({{ wsStore.latency }}ms)</span>
+        </span>
+      </n-space>
+
+      <n-button v-if="wsStore.status !== 'connected'" text type="primary" size="small" @click="reconnect">
+        重新连接
+      </n-button>
+    </n-space>
+
+    <!-- 错误提示 -->
+    <n-alert v-if="errorMsg" type="error" closable style="margin-top: 12px" @close="errorMsg = ''">
+      {{ errorMsg }}
+    </n-alert>
+  </n-card>
+</template>
+
+<script setup lang="ts">
+import { computed, ref, onMounted, watch } from 'vue'
+import { useRouter } from 'vue-router'
+import { useMessage } from 'naive-ui'
+import {
+  PulseOutline as PulseIcon,
+  PlayOutline as PlayIcon,
+  StopOutline as StopIcon,
+  RefreshOutline as RefreshIcon
+} from '@vicons/ionicons5'
+import { useWebSocketStore } from '@/stores/websocket'
+import axios from 'axios'
+
+const message = useMessage()
+const router = useRouter()
+const wsStore = useWebSocketStore()
+
+// 本地状态
+const isLoading = ref(false)
+const errorMsg = ref('')
+
+// 从WebSocket获取核心状态（实时）
+const coreStatus = computed(() => wsStore.coreStatus)
+const isRunning = computed(() => coreStatus.value?.is_running ?? false)
+const hasError = computed(() => !!coreStatus.value?.error)
+
+// 状态颜色
+const getStatusColor = computed(() => {
+  if (isLoading.value) return '#f0a020'
+  if (isRunning.value) return '#18a058'
+  if (hasError.value) return '#d03050'
+  return '#808080'
+})
+
+// 格式化运行时间
+const formatUptime = (seconds: number) => {
+  if (!seconds) return '0s'
+  const hrs = Math.floor(seconds / 3600)
+  const mins = Math.floor((seconds % 3600) / 60)
+  const secs = Math.floor(seconds % 60)
+  if (hrs > 0) return `${hrs}h ${mins}m`
+  if (mins > 0) return `${mins}m ${secs}s`
+  return `${secs}s`
+}
+
+// 使用WebSocket启动核心（更快速，无HTTP开销）
+const handleStart = async () => {
+  errorMsg.value = ''
+  isLoading.value = true
+
+  try {
+    // 优先使用WebSocket，如果不连接则回退到HTTP
+    if (wsStore.status === 'connected') {
+      wsStore.controlCore('start')
+      // 等待WebSocket推送状态更新（最多5秒）
+      await waitForStatusChange(true, 5000)
+    } else {
+      // HTTP备用方案
+      const res = await axios.post('/api/core/start')
+      if (!res.data.success) {
+        throw new Error(res.data.error || '启动失败')
+      }
+    }
+  } catch (error: any) {
+    errorMsg.value = error.response?.data?.error || error.message || '启动失败'
+    message.error(errorMsg.value)
+  } finally {
+    isLoading.value = false
+  }
+}
+
+// 使用WebSocket停止核心
+const handleStop = async () => {
+  const confirmed = confirm('确定要停止 Yosuga 核心吗？这将中断所有正在进行的对话和任务。')
+  if (!confirmed) return
+
+  isLoading.value = true
+  try {
+    if (wsStore.status === 'connected') {
+      wsStore.controlCore('stop')
+      await waitForStatusChange(false, 5000)
+    } else {
+      const res = await axios.post('/api/core/stop')
+      if (!res.data.success) {
+        throw new Error(res.data.error || '停止失败')
+      }
+    }
+    message.success('Yosuga 核心已停止')
+  } catch (error: any) {
+    errorMsg.value = error.response?.data?.error || error.message || '停止失败'
+    message.error(errorMsg.value)
+  } finally {
+    isLoading.value = false
+  }
+}
+
+// 等待状态变化辅助函数
+const waitForStatusChange = (targetRunning: boolean, timeout: number): Promise<void> => {
+  return new Promise((resolve, reject) => {
+    const checkInterval = setInterval(() => {
+      if (coreStatus.value?.is_running === targetRunning) {
+        clearInterval(checkInterval)
+        clearTimeout(timeoutId)
+        resolve()
+      }
+    }, 100)
+
+    const timeoutId = setTimeout(() => {
+      clearInterval(checkInterval)
+      resolve() // 超时也不报错，依赖后续状态更新
+    }, timeout)
+  })
+}
+
+// 刷新状态（HTTP备用）
+const refreshStatus = async () => {
+  try {
+    const res = await axios.get('/api/core/status')
+    if (res.data.success) {
+      // 手动更新store（如果需要）
+      message.success('状态已刷新')
+    }
+  } catch (error) {
+    message.error('刷新状态失败')
+  }
+}
+
+// 重新连接WebSocket
+const reconnect = () => {
+  wsStore.disconnect()
+  setTimeout(() => wsStore.connect(), 500)
+}
+
+// 监听WebSocket连接状态
+watch(() => wsStore.status, (newStatus) => {
+  if (newStatus === 'disconnected' && isRunning.value) {
+    // 如果断开但核心在运行，尝试重新连接
+    setTimeout(() => wsStore.connect(), 1000)
+  }
+})
+
+onMounted(() => {
+  // 确保WebSocket连接
+  if (wsStore.status !== 'connected') {
+    wsStore.connect()
+  }
+})
+</script>
+
+<style scoped>
+.core-control {
+  background: linear-gradient(135deg, var(--n-card-color) 0%, var(--n-action-color) 100%);
+  border: 1px solid var(--n-border-color);
+}
+
+.core-title {
+  font-weight: 600;
+  font-size: 16px;
+  display: flex;
+  align-items: center;
+}
+
+.core-status {
+  font-size: 13px;
+  color: var(--n-text-color-3);
+  margin-top: 4px;
+}
+
+.error-text {
+  color: #d03050;
+  font-weight: 500;
+}
+
+:deep(.n-badge .n-badge-dot) {
+  width: 10px;
+  height: 10px;
+}
+</style>

src/server_view/frontend/src/main.ts

@@ -0,0 +1,13 @@
+import { createApp } from 'vue'
+import { createPinia } from 'pinia'
+import naive from 'naive-ui'
+import App from './App.vue'
+import router from './router'
+import { useDiagnosticsStore } from './stores/diagnostics'
+
+export { useDiagnosticsStore }
+const app = createApp(App)
+app.use(createPinia())
+app.use(router)
+app.use(naive)
+app.mount('#app')

src/server_view/frontend/src/router/index.ts

@@ -0,0 +1,46 @@
+import { createRouter, createWebHistory } from 'vue-router'
+import Dashboard from '@/views/Dashboard.vue'
+import Monitor from '@/views/Monitor.vue'
+import Devices from '@/views/Devices.vue'
+import Config from '@/views/Config.vue'
+import Logs from '@/views/Logs.vue'
+
+const routes = [
+  {
+    path: '/',
+    name: 'dashboard',
+    component: Dashboard,
+    meta: { title: '仪表板' }
+  },
+  {
+    path: '/monitor',
+    name: 'monitor',
+    component: Monitor,
+    meta: { title: '系统监控' }
+  },
+  {
+    path: '/devices',
+    name: 'devices',
+    component: Devices,
+    meta: { title: '设备管理' }
+  },
+  {
+    path: '/config',
+    name: 'config',
+    component: Config,
+    meta: { title: '配置管理' }
+  },
+  {
+    path: '/logs',
+    name: 'logs',
+    component: Logs,
+    meta: { title: '日志查看' }
+  }
+]
+
+const router = createRouter({
+  history: createWebHistory(),
+  routes
+})
+
+export default router

src/server_view/frontend/src/stores/config.ts

@@ -0,0 +1,95 @@
+import { defineStore } from 'pinia'
+import { ref } from 'vue'
+import axios from 'axios'
+
+export interface ASRConfig {
+  enabled: boolean
+  url: string
+  model_name: string
+}
+
+export interface TTSConfig {
+  enabled: boolean
+  host: string
+  port: number
+  gpt_model_name: string
+  sovits_model_name: string
+  streaming: boolean
+}
+
+export interface AIConfig {
+  api_key: string | null
+  base_url: string
+  model_name: string
+  temperature: number
+  max_tokens: number
+}
+
+export interface AppConfigState {
+  version: string
+  debug: boolean
+  ai: AIConfig
+  tts: TTSConfig
+  asr: ASRConfig
+  auto_agent: any
+  llm_core: any
+}
+
+export const useConfigStore = defineStore('config', () => {
+  const config = ref<AppConfigState | null>(null)
+  const loading = ref(false)
+  const saving = ref(false)
+
+  const fetchConfig = async () => {
+    loading.value = true
+    try {
+      const response = await axios.get('/api/config')
+      if (response.data.success) {
+        config.value = response.data.data
+      }
+    } catch (error) {
+      console.error('获取配置失败:', error)
+      throw error
+    } finally {
+      loading.value = false
+    }
+  }
+
+  const updateConfig = async (section: string, data: any) => {
+    saving.value = true
+    try {
+      const response = await axios.post(`/api/config/${section}`, data)
+      if (response.data.success) {
+        if (config.value) {
+          (config.value as any)[section] = data
+        }
+        return true
+      }
+      return false
+    } catch (error) {
+      console.error('更新配置失败:', error)
+      throw error
+    } finally {
+      saving.value = false
+    }
+  }
+
+  const reloadConfig = async () => {
+    try {
+      const response = await axios.post('/api/config/reload')
+      return response.data.success
+    } catch (error) {
+      console.error('重载配置失败:', error)
+      throw error
+    }
+  }
+
+  return {
+    config,
+    loading,
+    saving,
+    fetchConfig,
+    updateConfig,
+    reloadConfig
+  }
+})

src/server_view/frontend/src/stores/core.ts

@@ -0,0 +1,41 @@
+// core.ts - 现在仅作为HTTP备用，主要逻辑迁移到websocket.ts
+import { defineStore } from 'pinia'
+import { ref } from 'vue'
+import axios from 'axios'
+
+export const useCoreStore = defineStore('core', () => {
+  // 保留HTTP方法作为备用
+  const fetchStatus = async () => {
+    try {
+      const response = await axios.get('/api/core/status')
+      return response.data.data
+    } catch (error) {
+      console.error('获取核心状态失败:', error)
+      return null
+    }
+  }
+
+  const startCore = async (): Promise<boolean> => {
+    try {
+      const response = await axios.post('/api/core/start')
+      return response.data.success
+    } catch (error) {
+      return false
+    }
+  }
+
+  const stopCore = async (): Promise<boolean> => {
+    try {
+      const response = await axios.post('/api/core/stop')
+      return response.data.success
+    } catch (error) {
+      return false
+    }
+  }
+
+  return {
+    fetchStatus,
+    startCore,
+    stopCore
+  }
+})

src/server_view/frontend/src/stores/diagnostics.ts

@@ -0,0 +1,199 @@
+import { defineStore } from 'pinia'
+import { ref, computed } from 'vue'
+import axios from 'axios'
+import { useWebSocketStore } from './websocket'
+
+export interface CheckResult {
+  name: string
+  status: 'healthy' | 'unhealthy' | 'unknown' | 'checking'
+  message: string
+  details: {
+    host?: string
+    port?: number
+    error?: string
+    tcp_ok?: boolean
+  }
+  latency_ms: number
+  timestamp: number
+}
+
+export interface DiagnosticsReport {
+  overall_status: 'healthy' | 'unhealthy' | 'unknown'
+  checks: CheckResult[]
+  summary: {
+    healthy: number
+    unhealthy: number
+    unknown: number
+    total: number
+  }
+  generated_at: number
+  version: string
+}
+
+export interface ModuleHealth {
+  module: string
+  status: 'healthy' | 'unhealthy' | 'unknown' | 'checking'
+  message: string
+  details: {
+    host?: string
+    port?: number
+    error?: string
+  }
+  latency_ms?: number
+  lastCheck: number
+}
+
+export const useDiagnosticsStore = defineStore('diagnostics', () => {
+  const wsStore = useWebSocketStore()
+
+  const isRunningFullCheck = ref(false)
+  const lastReport = ref<DiagnosticsReport | null>(null)
+  const moduleHealth = ref<Record<string, ModuleHealth>>({
+    asr: { module: 'asr', status: 'unknown', message: '未检查', details: {}, lastCheck: 0 },
+    tts: { module: 'tts', status: 'unknown', message: '未检查', details: {}, lastCheck: 0 },
+    ai: { module: 'ai', status: 'unknown', message: '未检查', details: {}, lastCheck: 0 },
+    auto_agent: { module: 'auto_agent', status: 'unknown', message: '未检查', details: {}, lastCheck: 0 },
+    llm_core: { module: 'llm_core', status: 'unknown', message: '未检查', details: {}, lastCheck: 0 }
+  })
+
+  const healthyCount = computed(() =>
+    Object.values(moduleHealth.value).filter(m => m.status === 'healthy').length
+  )
+
+  const hasUnhealthy = computed(() =>
+    Object.values(moduleHealth.value).some(m => m.status === 'unhealthy')
+  )
+
+  const overallStatus = computed(() => {
+    if (hasUnhealthy.value) return 'error'
+    if (healthyCount.value === 5) return 'success'
+    return 'warning'
+  })
+
+  const runFullDiagnostics = async (): Promise<DiagnosticsReport> => {
+    isRunningFullCheck.value = true
+    try {
+      const response = await axios.post<DiagnosticsReport>('/api/diagnostics/run')
+      lastReport.value = response.data
+
+      response.data.checks.forEach(check => {
+        const moduleMap: Record<string, string> = {
+          'ASR服务': 'asr',
+          'TTS服务': 'tts',
+          'AI服务': 'ai',
+          '自动代理服务': 'auto_agent',
+          'LLM核心': 'llm_core'
+        }
+
+        const moduleKey = moduleMap[check.name]
+        if (moduleKey && moduleHealth.value[moduleKey]) {
+          moduleHealth.value[moduleKey] = {
+            module: moduleKey,
+            status: check.status,
+            message: check.message,
+            details: check.details,
+            latency_ms: check.latency_ms,
+            lastCheck: Date.now()
+          }
+        }
+      })
+
+      return response.data
+    } finally {
+      isRunningFullCheck.value = false
+    }
+  }
+
+  const checkModule = async (module: string) => {
+    if (moduleHealth.value[module]) {
+      moduleHealth.value[module].status = 'checking'
+    }
+
+    try {
+      if (wsStore.status === 'connected') {
+        wsStore.socket?.emit('check_module_health', { module })
+        return
+      }
+
+      const response = await axios.get(`/api/diagnostics/check/${module}`)
+      if (response.data.success) {
+        updateModuleHealth(module, response.data.data)
+      }
+    } catch (error) {
+      moduleHealth.value[module] = {
+        module,
+        status: 'unhealthy',
+        message: '检查失败',
+        details: { error: '网络请求失败' },
+        lastCheck: Date.now()
+      }
+    }
+  }
+
+  const updateModuleHealth = (module: string, data: CheckResult) => {
+    moduleHealth.value[module] = {
+      module,
+      status: data.status,
+      message: data.message,
+      details: data.details,
+      latency_ms: data.latency_ms,
+      lastCheck: Date.now()
+    }
+  }
+
+  const checkAllModules = async () => {
+    const modules = ['asr', 'tts', 'ai', 'auto_agent', 'llm_core']
+    await Promise.all(modules.map(m => checkModule(m)))
+  }
+
+  const initWebSocketListeners = () => {
+    wsStore.socket?.on('module_health_result', (result: {
+      success: boolean
+      module?: string
+      data?: CheckResult
+      error?: string
+    }) => {
+      if (result.success && result.module && result.data) {
+        updateModuleHealth(result.module, result.data)
+      }
+    })
+
+    wsStore.socket?.on('module_status_update', (update: {
+      module: string
+      status: CheckResult
+    }) => {
+      updateModuleHealth(update.module, update.status)
+    })
+  }
+
+  let autoCheckInterval: NodeJS.Timeout | null = null
+
+  const startAutoCheck = (intervalMs = 30000) => {
+    stopAutoCheck()
+    autoCheckInterval = setInterval(() => {
+      checkAllModules()
+    }, intervalMs)
+  }
+
+  const stopAutoCheck = () => {
+    if (autoCheckInterval) {
+      clearInterval(autoCheckInterval)
+      autoCheckInterval = null
+    }
+  }
+
+  return {
+    isRunningFullCheck,
+    lastReport,
+    moduleHealth,
+    healthyCount,
+    hasUnhealthy,
+    overallStatus,
+    runFullDiagnostics,
+    checkModule,
+    checkAllModules,
+    initWebSocketListeners,
+    startAutoCheck,
+    stopAutoCheck
+  }
+})

src/server_view/frontend/src/stores/preferences.ts

@@ -0,0 +1,117 @@
+import { defineStore } from 'pinia'
+import { ref, watch } from 'vue'
+import axios from 'axios'
+
+export const usePreferencesStore = defineStore('preferences', () => {
+  // 用户偏好状态
+  const preferences = ref({
+    theme: 'light',           // 'light' | 'dark' | 'auto'
+    sidebarCollapsed: false,
+    logLevel: 'ALL',          // 默认日志等级筛选
+    logAutoScroll: true,      // 日志自动滚动
+    refreshInterval: 2000,    // 数据刷新间隔(ms)
+    lastVisitedPage: 'dashboard',
+    favoriteConfigs: [] as string[], // 收藏的配置项
+    notifications: {
+      coreStatus: true,       // 核心状态变化通知
+      errors: true            // 错误通知
+    }
+  })
+
+  const isLoaded = ref(false)
+
+  // 从服务器加载偏好
+  const loadPreferences = async () => {
+    try {
+      const response = await axios.get('/api/preferences')
+      if (response.data.success) {
+        preferences.value = { ...preferences.value, ...response.data.data }
+        applyTheme(preferences.value.theme)
+      }
+    } catch (error) {
+      console.error('加载偏好失败:', error)
+      // 从 localStorage 回退
+      const local = localStorage.getItem('yosuga_preferences')
+      if (local) {
+        preferences.value = { ...preferences.value, ...JSON.parse(local) }
+      }
+    } finally {
+      isLoaded.value = true
+    }
+  }
+
+  // 保存偏好到服务器和本地
+  const savePreferences = async () => {
+    try {
+      await axios.post('/api/preferences', preferences.value)
+      localStorage.setItem('yosuga_preferences', JSON.stringify(preferences.value))
+    } catch (error) {
+      console.error('保存偏好失败:', error)
+      // 至少保存到本地
+      localStorage.setItem('yosuga_preferences', JSON.stringify(preferences.value))
+    }
+  }
+
+  // 应用主题
+  const applyTheme = (theme: string) => {
+    if (theme === 'dark') {
+      document.documentElement.classList.add('dark')
+    } else if (theme === 'light') {
+      document.documentElement.classList.remove('dark')
+    } else {
+      // auto: 跟随系统
+      if (window.matchMedia('(prefers-color-scheme: dark)').matches) {
+        document.documentElement.classList.add('dark')
+      } else {
+        document.documentElement.classList.remove('dark')
+      }
+    }
+  }
+
+  // 设置主题
+  const setTheme = (theme: string) => {
+    preferences.value.theme = theme
+    applyTheme(theme)
+    savePreferences()
+  }
+
+  // 切换侧边栏
+  const toggleSidebar = () => {
+    preferences.value.sidebarCollapsed = !preferences.value.sidebarCollapsed
+    savePreferences()
+  }
+
+  // 设置日志等级
+  const setLogLevel = (level: string) => {
+    preferences.value.logLevel = level
+    savePreferences()
+  }
+
+  // 监听系统主题变化
+  if (window.matchMedia) {
+    window.matchMedia('(prefers-color-scheme: dark)').addEventListener('change', (e) => {
+      if (preferences.value.theme === 'auto') {
+        applyTheme('auto')
+      }
+    })
+  }
+
+  // 自动保存（防抖）
+  let saveTimer: NodeJS.Timeout
+  watch(preferences, () => {
+    clearTimeout(saveTimer)
+    saveTimer = setTimeout(() => {
+      savePreferences()
+    }, 500)
+  }, { deep: true })
+
+  return {
+    preferences,
+    isLoaded,
+    loadPreferences,
+    savePreferences,
+    setTheme,
+    toggleSidebar,
+    setLogLevel
+  }
+})

src/server_view/frontend/src/stores/websocket.ts

@@ -0,0 +1,156 @@
+import { defineStore } from 'pinia'
+import { ref, computed } from 'vue'
+import { io, Socket } from 'socket.io-client'
+
+// 系统状态类型定义
+export interface SystemStats {
+  cpu: { percent: number; count: number }
+  memory: {
+    total: number; available: number;
+    percent: number; used: number; free: number
+  }
+  disk: { total: number; used: number; free: number; percent: number }
+  process: {
+    memory_percent: number; cpu_percent: number;
+    threads: number; uptime: number
+  }
+  timestamp: number
+}
+
+export interface CoreStatus {
+  is_running: boolean
+  pid: number
+  uptime: number
+  error: string | null
+  thread_alive: boolean
+  start_time: string | null
+}
+
+export const useWebSocketStore = defineStore('websocket', () => {
+  const socket = ref<Socket | null>(null)
+  const status = ref<'disconnected' | 'connecting' | 'connected'>('disconnected')
+  const logs = ref<string[]>([])
+  const currentFilter = ref('ALL')
+
+  // 新增：系统状态（来自WebSocket实时推送）
+  const systemStats = ref<SystemStats | null>(null)
+  const coreStatus = ref<CoreStatus | null>(null)
+
+  // 连接状态时间戳
+  const lastPing = ref<number>(0)
+  const latency = computed(() => {
+    if (!lastPing.value) return 0
+    return Date.now() - lastPing.value
+  })
+
+  const connect = () => {
+    if (socket.value?.connected) return
+
+    status.value = 'connecting'
+    socket.value = io('http://localhost:8089', {
+      transports: ['websocket'],
+      autoConnect: true,
+      reconnection: true,
+      reconnectionDelay: 1000,
+      reconnectionAttempts: 5,
+    })
+
+    socket.value.on('connect', () => {
+      status.value = 'connected'
+      lastPing.value = Date.now()
+      // 连接时自动订阅日志
+      subscribeWithLevel(currentFilter.value)
+    })
+
+    socket.value.on('disconnect', (reason) => {
+      status.value = 'disconnected'
+      console.log(`WebSocket断开: ${reason}`)
+    })
+
+    socket.value.on('connect_error', (err) => {
+      console.error('WebSocket连接错误:', err)
+      status.value = 'disconnected'
+    })
+
+    // 日志接收
+    socket.value.on('log_line', (data: { line: string; timestamp?: string; level?: string }) => {
+      logs.value.push(data.line)
+      if (logs.value.length > 1000) {
+        logs.value.shift()
+      }
+    })
+
+    // 新增：系统状态实时推送
+    socket.value.on('system_stats', (response: { success: boolean; data: SystemStats }) => {
+      if (response.success && response.data) {
+        systemStats.value = response.data
+        lastPing.value = Date.now()
+      }
+    })
+
+    // 新增：核心状态实时推送（替代HTTP轮询）
+    socket.value.on('core_status', (response: { success: boolean; data: CoreStatus; message?: string }) => {
+      if (response.success && response.data) {
+        coreStatus.value = response.data
+        console.log('核心状态更新:', response.data.is_running ? '运行中' : '已停止', response.message || '')
+      }
+    })
+
+    // 控制操作结果回调
+    socket.value.on('core_control_result', (result: { success: boolean; error?: string; message?: string; data?: any }) => {
+      if (!result.success) {
+        console.error('核心控制失败:', result.error)
+      } else {
+        console.log('核心控制成功:', result.message)
+      }
+    })
+  }
+
+  // 新增：通过WebSocket控制核心（替代HTTP请求）
+  const controlCore = (action: 'start' | 'stop') => {
+    socket.value?.emit('control_core', { action })
+  }
+
+  const subscribeWithLevel = (level: string) => {
+    currentFilter.value = level
+    socket.value?.emit('subscribe_logs', { level })
+  }
+
+  const disconnect = () => {
+    socket.value?.disconnect()
+  }
+
+  const clearLogs = () => {
+    logs.value = []
+  }
+
+  // 心跳检测
+  const startHeartbeat = () => {
+    const interval = setInterval(() => {
+      if (status.value === 'connected' && socket.value) {
+        // 如果超过10秒没有收到system_stats，认为连接可能有问题
+        if (lastPing.value && (Date.now() - lastPing.value > 10000)) {
+          console.warn('WebSocket可能卡顿，尝试重新连接...')
+          socket.value.connect()
+        }
+      }
+    }, 5000)
+    return () => clearInterval(interval)
+  }
+
+  return {
+    socket,
+    status,
+    logs,
+    currentFilter,
+    systemStats,      // 导出系统状态
+    coreStatus,       // 导出核心状态
+    latency,
+    connect,
+    disconnect,
+    clearLogs,
+    subscribeWithLevel,
+    controlCore,      // 导出控制函数
+    startHeartbeat
+  }
+})

src/server_view/frontend/src/views/Config.vue

@@ -0,0 +1,554 @@
+<template>
+  <div class="config-page">
+    <n-card title="全局配置管理" :bordered="false">
+      <template #header-extra>
+        <n-space>
+          <n-button @click="fetchConfig" :loading="loading">
+            <template #icon><n-icon><refresh-icon /></n-icon></template>
+            刷新
+          </n-button>
+          <n-button type="primary" @click="handleSave" :loading="saving">
+            <template #icon><n-icon><save-icon /></n-icon></template>
+            保存更改
+          </n-button>
+          <n-button @click="showJson = true">
+            <template #icon><n-icon><code-icon /></n-icon></template>
+            查看JSON
+          </n-button>
+        </n-space>
+      </template>
+
+      <n-tabs type="line" animated v-model:value="activeTab">
+        <!-- AI 模型配置 -->
+        <n-tab-pane name="ai" tab="AI 模型">
+          <n-alert type="info" style="margin-bottom: 16px" :show-icon="false">
+            配置主对话AI服务，支持OpenAI、DeepSeek或本地LM Studio
+          </n-alert>
+
+          <n-form
+            v-if="formData.ai"
+            :model="formData.ai"
+            label-placement="left"
+            label-width="140px"
+            require-mark-placement="right-hanging"
+          >
+            <n-form-item label="API Key">
+              <n-input
+                v-model:value="formData.ai.api_key"
+                type="password"
+                placeholder="sk-xxxxxxxxxxxxxxxx"
+                show-password-on="click"
+              />
+              <template #feedback>留空使用本地模型或无需认证的端点</template>
+            </n-form-item>
+
+            <n-form-item label="Base URL">
+              <n-input
+                v-model:value="formData.ai.base_url"
+                placeholder="http://localhost:1234/v1"
+              />
+              <template #feedback>API端点地址，本地LM Studio通常为 http://localhost:1234/v1</template>
+            </n-form-item>
+
+            <n-form-item label="模型名称">
+              <n-input
+                v-model:value="formData.ai.model_name"
+                placeholder="qwen/qwen3-4b-2507"
+              />
+              <template #feedback>例如: gpt-4, deepseek-chat, qwen/qwen3-4b-2507</template>
+            </n-form-item>
+
+            <n-form-item label="Temperature">
+              <n-space align="center" style="width: 100%">
+                <n-slider
+                  v-model:value="formData.ai.temperature"
+                  :min="0"
+                  :max="2"
+                  :step="0.1"
+                  style="width: 200px"
+                />
+                <n-input-number
+                  v-model:value="formData.ai.temperature"
+                  :min="0"
+                  :max="2"
+                  :step="0.1"
+                  style="width: 100px"
+                />
+              </n-space>
+              <template #feedback>控制随机性: 0=确定性, 1=平衡, 2=创造性</template>
+            </n-form-item>
+
+            <n-form-item label="Max Tokens">
+              <n-input-number
+                v-model:value="formData.ai.max_tokens"
+                :min="256"
+                :max="32768"
+                style="width: 200px"
+              />
+              <template #feedback>单次回复最大token数，通常设为8192</template>
+            </n-form-item>
+
+            <n-form-item label="超时时间(秒)">
+              <n-input-number
+                v-model:value="formData.ai.timeout"
+                :min="5"
+                :max="300"
+                style="width: 200px"
+              />
+              <template #feedback>API请求超时时间</template>
+            </n-form-item>
+          </n-form>
+        </n-tab-pane>
+
+        <!-- TTS 语音合成配置 -->
+        <n-tab-pane name="tts" tab="语音合成 (TTS)">
+          <n-alert type="info" style="margin-bottom: 16px" :show-icon="false">
+            GPT-SoVITS 语音合成服务配置
+          </n-alert>
+
+          <n-form
+            v-if="formData.tts"
+            :model="formData.tts"
+            label-placement="left"
+            label-width="140px"
+          >
+            <n-form-item label="启用 TTS">
+              <n-switch v-model:value="formData.tts.enabled" />
+            </n-form-item>
+
+            <n-divider title-placement="left">服务端连接</n-divider>
+
+            <n-form-item label="主机地址">
+              <n-input
+                v-model:value="formData.tts.host"
+                placeholder="localhost"
+              />
+            </n-form-item>
+
+            <n-form-item label="端口">
+              <n-input-number
+                v-model:value="formData.tts.port"
+                :min="1"
+                :max="65535"
+                style="width: 200px"
+              />
+              <template #feedback>GPT-SoVITS API默认端口 9880 或 20261</template>
+            </n-form-item>
+
+            <n-form-item label="API Key">
+              <n-input
+                v-model:value="formData.tts.api_key"
+                type="password"
+                placeholder="可选"
+                show-password-on="click"
+              />
+            </n-form-item>
+
+            <n-divider title-placement="left">模型配置</n-divider>
+
+            <n-form-item label="GPT模型路径">
+              <n-input
+                v-model:value="formData.tts.gpt_model_name"
+                placeholder="GPT_weights_v2Pro/Yosuga_Airi-e32.ckpt"
+              />
+              <template #feedback>相对于项目根目录的路径</template>
+            </n-form-item>
+
+            <n-form-item label="SoVITS模型路径">
+              <n-input
+                v-model:value="formData.tts.sovits_model_name"
+                placeholder="SoVITS_weights_v2Pro/Yosuga_Airi_e16_s864.pth"
+              />
+              <template #feedback>相对于项目根目录的路径</template>
+            </n-form-item>
+
+            <n-divider title-placement="left">音频设置</n-divider>
+
+            <n-form-item label="参考音频路径">
+              <n-input
+                v-model:value="formData.tts.reference_audio"
+                placeholder="./using/reference.wav"
+              />
+              <template #feedback>用于声音克隆的参考音频文件路径</template>
+            </n-form-item>
+
+            <n-form-item label="流式输出">
+              <n-switch v-model:value="formData.tts.streaming" />
+              <template #feedback>启用流式传输可降低延迟</template>
+            </n-form-item>
+
+            <n-form-item label="语速倍率">
+              <n-space align="center" style="width: 100%">
+                <n-slider
+                  v-model:value="formData.tts.speed"
+                  :min="0.6"
+                  :max="1.65"
+                  :step="0.05"
+                  style="width: 200px"
+                />
+                <span style="min-width: 50px">{{ formData.tts.speed }}x</span>
+              </n-space>
+            </n-form-item>
+          </n-form>
+        </n-tab-pane>
+
+        <!-- ASR 语音识别配置 -->
+        <n-tab-pane name="asr" tab="语音识别 (ASR)">
+          <n-alert type="info" style="margin-bottom: 16px" :show-icon="false">
+            Faster-Whisper 语音识别服务配置
+          </n-alert>
+
+          <n-form
+            v-if="formData.asr"
+            :model="formData.asr"
+            label-placement="left"
+            label-width="140px"
+          >
+            <n-form-item label="启用 ASR">
+              <n-switch v-model:value="formData.asr.enabled" />
+            </n-form-item>
+
+            <n-form-item label="服务 URL">
+              <n-input
+                v-model:value="formData.asr.url"
+                placeholder="http://localhost:20260/"
+              />
+              <template #feedback>ASR API端点地址，默认端口20260</template>
+            </n-form-item>
+
+            <n-form-item label="模型名称">
+              <n-input
+                v-model:value="formData.asr.model_name"
+                placeholder="fast-whisper"
+              />
+              <template #feedback>使用的ASR模型标识</template>
+            </n-form-item>
+
+            <n-form-item label="API Key">
+              <n-input
+                v-model:value="formData.asr.api_key"
+                type="password"
+                placeholder="可选"
+                show-password-on="click"
+              />
+            </n-form-item>
+          </n-form>
+        </n-tab-pane>
+
+        <!-- 自动代理配置 -->
+        <n-tab-pane name="auto_agent" tab="自动代理 (Agent)">
+          <n-alert type="info" style="margin-bottom: 16px" :show-icon="false">
+            UI-TARS 自动化操作代理配置，用于GUI控制
+          </n-alert>
+
+          <n-form
+            v-if="formData.auto_agent"
+            :model="formData.auto_agent"
+            label-placement="left"
+            label-width="140px"
+          >
+            <n-form-item label="启用自动代理">
+              <n-switch v-model:value="formData.auto_agent.enabled" />
+            </n-form-item>
+
+            <n-form-item label="部署类型">
+              <n-select
+                v-model:value="formData.auto_agent.deployment_type"
+                :options="[
+                  { label: 'LM Studio (本地)', value: 'lmstudio' },
+                  { label: 'vLLM (本地服务器)', value: 'vllm' },
+                  { label: 'Ollama (本地)', value: 'ollama' },
+                  { label: '云端 API', value: 'cloud' }
+                ]"
+                style="width: 200px"
+              />
+            </n-form-item>
+
+            <n-form-item label="Base URL">
+              <n-input
+                v-model:value="formData.auto_agent.base_url"
+                placeholder="http://localhost:1234/v1"
+              />
+              <template #feedback>UI-TARS模型服务端点</template>
+            </n-form-item>
+
+            <n-form-item label="模型名称">
+              <n-input
+                v-model:value="formData.auto_agent.model_name"
+                placeholder="ui-tars-1.5-7b@q4_k_m"
+              />
+              <template #feedback>例如: ui-tars-1.5-7b, ui-tars-2.0-7b-sft</template>
+            </n-form-item>
+
+            <n-form-item label="API Key">
+              <n-input
+                v-model:value="formData.auto_agent.api_key"
+                type="password"
+                placeholder="可选，云端服务时需要"
+                show-password-on="click"
+              />
+            </n-form-item>
+
+            <n-form-item label="Temperature">
+              <n-space align="center" style="width: 100%">
+                <n-slider
+                  v-model:value="formData.auto_agent.temperature"
+                  :min="0"
+                  :max="1"
+                  :step="0.05"
+                  style="width: 200px"
+                />
+                <n-input-number
+                  v-model:value="formData.auto_agent.temperature"
+                  :min="0"
+                  :max="1"
+                  :step="0.05"
+                  style="width: 100px"
+                />
+              </n-space>
+              <template #feedback>UI任务建议低温度(0.1-0.3)提高准确性</template>
+            </n-form-item>
+
+            <n-form-item label="Max Tokens">
+              <n-input-number
+                v-model:value="formData.auto_agent.max_tokens"
+                :min="2048"
+                :max="128000"
+                :step="1024"
+                style="width: 200px"
+              />
+              <template #feedback>UI-TARS需要较大token数来处理截图，建议16384+</template>
+            </n-form-item>
+          </n-form>
+        </n-tab-pane>
+
+        <!-- LLM 核心配置 -->
+        <n-tab-pane name="llm_core" tab="LLM 核心">
+          <n-alert type="info" style="margin-bottom: 16px" :show-icon="false">
+            Yosuga LLM Core 行为配置
+          </n-alert>
+
+          <n-form
+            v-if="formData.llm_core"
+            :model="formData.llm_core"
+            label-placement="left"
+            label-width="140px"
+          >
+            <n-form-item label="启用 LLM核心">
+              <n-switch v-model:value="formData.llm_core.enabled" />
+            </n-form-item>
+
+            <n-form-item label="启用历史记录">
+              <n-switch v-model:value="formData.llm_core.enable_history" />
+              <template #feedback>开启后LLM会记住对话上下文</template>
+            </n-form-item>
+
+            <n-form-item label="最大上下文 Tokens">
+              <n-input-number
+                v-model:value="formData.llm_core.max_context_tokens"
+                :min="512"
+                :max="32768"
+                :step="512"
+                style="width: 200px"
+              />
+              <template #feedback>上下文长度限制，超出会触发记忆清理</template>
+            </n-form-item>
+
+            <n-form-item label="回复语言">
+              <n-select
+                v-model:value="formData.llm_core.language"
+                :options="[
+                  { label: '日本语 (Japanese)', value: '日本语' },
+                  { label: '中文 (Chinese)', value: 'zh_CN' },
+                  { label: 'English', value: 'en' }
+                ]"
+                style="width: 200px"
+              />
+            </n-form-item>
+
+            <n-form-item label="角色设定">
+              <n-input
+                v-model:value="formData.llm_core.role_character"
+                type="textarea"
+                :rows="6"
+                placeholder="输入角色设定..."
+              />
+              <template #feedback>
+                定义AI助手的性格、背景和行为方式
+                <n-button text type="primary" size="tiny" @click="resetRoleToDefault">
+                  恢复默认
+                </n-button>
+              </template>
+            </n-form-item>
+          </n-form>
+        </n-tab-pane>
+      </n-tabs>
+    </n-card>
+
+    <!-- JSON预览模态框 -->
+    <n-modal v-model:show="showJson" title="当前配置 (JSON)" style="width: 800px">
+      <n-card>
+        <n-code :code="jsonPreview" language="json" style="max-height: 500px; overflow: auto;" />
+        <n-space justify="end" style="margin-top: 16px;">
+          <n-button @click="copyJson">复制到剪贴板</n-button>
+          <n-button @click="showJson = false">关闭</n-button>
+        </n-space>
+      </n-card>
+    </n-modal>
+  </div>
+</template>
+
+<script setup lang="ts">
+import { ref, onMounted, reactive, computed } from 'vue'
+import { useMessage } from 'naive-ui'
+import {
+  RefreshOutline as RefreshIcon,
+  SaveOutline as SaveIcon,
+  CodeWorkingOutline as CodeIcon
+} from '@vicons/ionicons5'
+import axios from 'axios'
+
+const message = useMessage()
+const activeTab = ref('ai')
+const loading = ref(false)
+const saving = ref(false)
+const showJson = ref(false)
+
+// 完整的配置数据结构（与后端config.py完全对应）
+const formData = reactive({
+  ai: {
+    api_key: '',
+    base_url: '',
+    model_name: '',
+    timeout: 30,
+    temperature: 0.7,
+    max_tokens: 8192
+  },
+  tts: {
+    enabled: true,
+    api_key: null as string | null,
+    gpt_model_name: '',
+    sovits_model_name: '',
+    host: 'localhost',
+    port: 20261,
+    reference_audio: './using/reference.wav',
+    streaming: true,
+    speed: 1.0
+  },
+  asr: {
+    enabled: true,
+    api_key: null as string | null,
+    model_name: 'fast-whisper',
+    url: 'http://localhost:20260/'
+  },
+  auto_agent: {
+    enabled: true,
+    api_key: null as string | null,
+    deployment_type: 'lmstudio',
+    model_name: 'ui-tars-1.5-7b@q4_k_m',
+    base_url: 'http://localhost:1234/v1',
+    temperature: 0.1,
+    max_tokens: 16384
+  },
+  llm_core: {
+    enabled: true,
+    role_character: '你是由Misakiotoha开发的助手稲葉愛理ちゃん，可以和用户一起玩游戏，聊天，做各种事情，性格抽象，没事爱整整活。',
+    max_context_tokens: 2048,
+    enable_history: true,
+    language: '日本语'
+  }
+})
+
+const defaultRole = '你是由Misakiotoha开发的助手稲葉愛理ちゃん，可以和用户一起玩游戏，聊天，做各种事情，性格抽象，没事爱整整活。'
+
+const jsonPreview = computed(() => {
+  return JSON.stringify({
+    ai: formData.ai,
+    tts: formData.tts,
+    asr: formData.asr,
+    auto_agent: formData.auto_agent,
+    llm_core: formData.llm_core
+  }, null, 2)
+})
+
+const fetchConfig = async () => {
+  loading.value = true
+  try {
+    const response = await axios.get('/api/config')
+    if (response.data.success) {
+      const cfg = response.data.data
+
+      // 深度合并配置（确保新字段也能显示）
+      Object.assign(formData.ai, cfg.ai || {})
+      Object.assign(formData.tts, cfg.tts || {})
+      Object.assign(formData.asr, cfg.asr || {})
+      Object.assign(formData.auto_agent, cfg.auto_agent || {})
+      Object.assign(formData.llm_core, cfg.llm_core || {})
+    }
+  } catch (error) {
+    console.error('获取配置失败:', error)
+    message.error('获取配置失败')
+  } finally {
+    loading.value = false
+  }
+}
+
+const handleSave = async () => {
+  saving.value = true
+  try {
+    // 保存当前选中的tab对应的配置节
+    const section = activeTab.value
+    const data = formData[section as keyof typeof formData]
+
+    const response = await axios.post(`/api/config/${section}`, data)
+
+    if (response.data.success) {
+      message.success('配置已保存并生效')
+    } else {
+      message.error('保存失败: ' + (response.data.message || '未知错误'))
+    }
+  } catch (error: any) {
+    console.error('保存配置失败:', error)
+    message.error('保存失败: ' + (error.response?.data?.detail || error.message))
+  } finally {
+    saving.value = false
+  }
+}
+
+const resetRoleToDefault = () => {
+  formData.llm_core.role_character = defaultRole
+  message.info('已恢复默认角色设定')
+}
+
+const copyJson = () => {
+  navigator.clipboard.writeText(jsonPreview.value)
+  message.success('已复制到剪贴板')
+}
+
+onMounted(() => {
+  fetchConfig()
+})
+</script>
+
+<style scoped>
+.config-page {
+  max-width: 1000px;
+  margin: 0 auto;
+}
+
+:deep(.n-form-item-feedback) {
+  font-size: 12px;
+  color: var(--n-text-color-3);
+  line-height: 1.4;
+}
+
+:deep(.n-divider) {
+  margin: 24px 0 16px 0;
+}
+
+:deep(.n-divider__title) {
+  font-size: 13px;
+  color: var(--n-text-color-3);
+  font-weight: 500;
+}
+</style>

src/server_view/frontend/src/views/Dashboard.vue

@@ -0,0 +1,560 @@
+<template>
+  <div class="dashboard">
+    <CoreControl style="margin-bottom: 16px" />
+
+    <!-- 系统体检报告卡片 -->
+    <n-card
+      v-if="lastReport"
+      :title="`系统体检报告 (v${lastReport.version})`"
+      class="diagnostics-card"
+      :bordered="false"
+      closable
+      @close="lastReport = null"
+    >
+      <n-space align="center" justify="space-between">
+        <n-space align="center">
+          <n-icon size="32" :color="statusColor">
+            <checkmark-circle-icon v-if="diagStore.overallStatus === 'success'" />
+            <warning-icon v-else-if="diagStore.overallStatus === 'warning'" />
+            <close-circle-icon v-else />
+          </n-icon>
+          <div>
+            <div :style="{ color: statusColor, fontWeight: 600 }">
+              {{ statusText }}
+            </div>
+            <div class="text-secondary">
+              {{ lastReport.summary.healthy }}/{{ lastReport.summary.total }} 项检查通过
+            </div>
+          </div>
+        </n-space>
+
+        <n-button text type="primary" @click="showDetails = true">
+          查看详情
+        </n-button>
+      </n-space>
+    </n-card>
+
+    <n-grid :x-gap="16" :y-gap="16" cols="1 s:2 l:4">
+      <n-gi>
+        <n-card class="stat-card" :bordered="false">
+          <n-statistic label="CPU 使用率" :value="systemStats?.cpu?.percent ?? 0">
+            <template #suffix>%</template>
+          </n-statistic>
+          <n-progress
+            type="line"
+            :percentage="systemStats?.cpu?.percent ?? 0"
+            :indicator-placement="'inside'"
+            :color="getProgressColor(systemStats?.cpu?.percent ?? 0)"
+          />
+        </n-card>
+      </n-gi>
+
+      <n-gi>
+        <n-card class="stat-card" :bordered="false">
+          <n-statistic label="内存使用" :value="memoryPercent">
+            <template #suffix>%</template>
+          </n-statistic>
+          <n-progress
+            type="line"
+            :percentage="memoryPercent"
+            :indicator-placement="'inside'"
+            :color="getProgressColor(memoryPercent)"
+          />
+        </n-card>
+      </n-gi>
+
+      <n-gi>
+        <n-card class="stat-card" :bordered="false">
+          <n-statistic label="运行时间" :value="uptimeFormatted" />
+          <div class="stat-detail">进程启动至今</div>
+        </n-card>
+      </n-gi>
+
+      <n-gi>
+        <n-card class="stat-card" :bordered="false">
+          <n-statistic label="服务健康" :value="`${diagStore.healthyCount}/5`" />
+          <div class="stat-detail">
+            <span :style="{ color: diagStore.hasUnhealthy ? '#d03050' : '#18a058' }">
+              {{ diagStore.hasUnhealthy ? '存在异常服务' : '所有服务正常' }}
+            </span>
+          </div>
+        </n-card>
+      </n-gi>
+    </n-grid>
+
+    <!-- 模块状态 - TCP端口检测 -->
+    <n-card title="模块状态" class="module-card" :bordered="false">
+      <template #header-extra>
+        <n-space>
+          <n-tag :type="checkingAll ? 'warning' : 'success'" size="small">
+            <n-spin v-if="checkingAll" :size="12" style="margin-right: 4px" />
+            {{ checkingAll ? '检测中...' : 'TCP端口检测' }}
+          </n-tag>
+          <n-button text type="primary" size="small" @click="refreshAllModules">
+            <template #icon><n-icon><refresh-icon /></n-icon></template>
+            刷新
+          </n-button>
+        </n-space>
+      </template>
+
+      <n-grid :x-gap="16" :y-gap="16" cols="2 s:3 l:5">
+        <n-gi v-for="(health, name) in moduleHealthDisplay" :key="name">
+          <div
+            class="module-item"
+            :class="{
+              active: health.status === 'healthy',
+              error: health.status === 'unhealthy',
+              checking: health.status === 'checking'
+            }"
+            @click="checkSingleModule(name)"
+          >
+            <n-icon size="24">
+              <checkmark-circle-icon v-if="health.status === 'healthy'" color="#18a058" />
+              <close-circle-icon v-else-if="health.status === 'unhealthy'" color="#d03050" />
+              <time-icon v-else-if="health.status === 'checking'" color="#f0a020" />
+              <help-circle-icon v-else color="#808080" />
+            </n-icon>
+
+            <div class="module-info">
+              <div class="module-name">{{ getModuleName(name) }}</div>
+
+              <n-tooltip>
+                <template #trigger>
+                  <div class="module-status" :class="health.status">
+                    <span v-if="health.status === 'healthy'">端口可连通</span>
+                    <span v-else-if="health.status === 'unhealthy'">端口不可达</span>
+                    <span v-else-if="health.status === 'checking'">检测中...</span>
+                    <span v-else>未检测</span>
+
+                    <span v-if="health.latency_ms && health.status === 'healthy'" class="latency">
+                      ({{ health.latency_ms < 1 ? '<1' : health.latency_ms.toFixed(0) }}ms)
+                    </span>
+                  </div>
+                </template>
+                <div style="font-size: 12px;">
+                  <div>主机: {{ health.details?.host || 'localhost' }}</div>
+                  <div>端口: {{ health.details?.port || getDefaultPort(name) }}</div>
+                  <div>上次检查: {{ formatTime(health.lastCheck) }}</div>
+                  <div v-if="health.details?.error" style="color: #ff4d4f;">
+                    错误: {{ health.details.error }}
+                  </div>
+                </div>
+              </n-tooltip>
+            </div>
+
+            <n-button
+              v-if="health.status === 'unhealthy'"
+              text
+              type="error"
+              size="tiny"
+              @click.stop="showModuleDetails(name)"
+            >
+              详情
+            </n-button>
+          </div>
+        </n-gi>
+      </n-grid>
+    </n-card>
+
+    <!-- 快速操作 -->
+    <n-card title="快速操作" class="action-card" :bordered="false">
+      <n-space>
+        <n-button
+          type="primary"
+          :loading="diagStore.isRunningFullCheck"
+          @click="runSystemDiagnostics"
+        >
+          <template #icon><n-icon><medical-icon /></n-icon></template>
+          {{ diagStore.isRunningFullCheck ? '体检中...' : '系统体检' }}
+        </n-button>
+
+        <n-button @click="handleReloadConfig">
+          <template #icon><n-icon><refresh-icon /></n-icon></template>
+          重载配置
+        </n-button>
+
+        <n-button @click="exportConfig">
+          <template #icon><n-icon><download-icon /></n-icon></template>
+          导出配置
+        </n-button>
+
+        <n-upload
+          action="/api/config/import"
+          accept=".json"
+          :show-file-list="false"
+          @finish="handleImportFinish"
+          @error="handleImportError"
+        >
+          <n-button>
+            <template #icon><n-icon><upload-icon /></n-icon></template>
+            导入配置
+          </n-button>
+        </n-upload>
+
+        <n-button @click="router.push('/logs')">
+          <template #icon><n-icon><terminal-icon /></n-icon></template>
+          查看日志
+        </n-button>
+      </n-space>
+    </n-card>
+
+    <!-- 体检详情抽屉 -->
+    <n-drawer v-model:show="showDetails" :width="600" placement="right">
+      <n-drawer-content title="详细体检报告">
+        <n-timeline v-if="lastReport">
+          <n-timeline-item
+            v-for="check in lastReport.checks"
+            :key="check.name"
+            :type="check.status === 'healthy' ? 'success' : check.status === 'unhealthy' ? 'error' : 'default'"
+            :title="check.name"
+            :content="check.message"
+            :time="`${check.latency_ms.toFixed(1)}ms`"
+          />
+        </n-timeline>
+      </n-drawer-content>
+    </n-drawer>
+
+    <!-- 模块详情Modal -->
+    <n-modal v-model:show="showModuleModal" :title="`${selectedModule} 连接详情`">
+      <n-card style="width: 450px" v-if="selectedModule">
+        <n-descriptions bordered :column="1" size="small">
+          <n-descriptions-item label="检测方式">
+            <n-tag size="small">TCP端口连通性</n-tag>
+          </n-descriptions-item>
+
+          <n-descriptions-item label="目标地址">
+            {{ currentModuleHealth?.details?.host || 'localhost' }}
+            :{{ currentModuleHealth?.details?.port || getDefaultPort(selectedModule) }}
+          </n-descriptions-item>
+
+          <n-descriptions-item label="连接状态">
+            <n-tag :type="currentModuleHealth?.status === 'healthy' ? 'success' : 'error'">
+              {{ currentModuleHealth?.status === 'healthy' ? '端口开放' : '端口关闭' }}
+            </n-tag>
+          </n-descriptions-item>
+
+          <n-descriptions-item label="响应延迟" v-if="currentModuleHealth?.latency_ms">
+            {{ currentModuleHealth.latency_ms.toFixed(2) }} ms
+          </n-descriptions-item>
+
+          <n-descriptions-item label="最后检测">
+            {{ formatTime(currentModuleHealth?.lastCheck || 0) }}
+          </n-descriptions-item>
+
+          <n-descriptions-item label="错误信息" v-if="currentModuleHealth?.details?.error">
+            <span style="color: #d03050;">
+              {{ currentModuleHealth.details.error }}
+            </span>
+          </n-descriptions-item>
+        </n-descriptions>
+
+        <n-alert
+          v-if="currentModuleHealth?.status === 'unhealthy'"
+          type="error"
+          style="margin-top: 16px"
+          :show-icon="true"
+        >
+          <template #header>排查建议</template>
+
+          <div v-if="currentModuleHealth?.details?.error === '连接被拒绝'" style="font-size: 13px;">
+            1. 服务可能已崩溃或未启动<br>
+            2. 检查防火墙是否放行端口<br>
+            3. 查看端口是否被其他进程占用
+          </div>
+          <div v-else-if="currentModuleHealth?.details?.error === '连接超时'" style="font-size: 13px;">
+            1. 服务未启动或网络不可达<br>
+            2. 检查IP地址和端口配置<br>
+            3. 确认服务监听的是0.0.0.0而非127.0.0.1
+          </div>
+          <div v-else style="font-size: 13px;">
+            1. 确认服务已启动<br>
+            2. 检查配置文件中的host/port<br>
+            3. 查看系统日志获取详细信息
+          </div>
+
+          <n-divider style="margin: 8px 0;" />
+          <n-space vertical size="small" style="font-size: 12px;">
+            <div>配置文件: <code>settings.json</code></div>
+            <div v-if="selectedModule === 'asr'">默认端口: 20260</div>
+            <div v-else-if="selectedModule === 'tts'">默认端口: 20261</div>
+            <div v-else-if="selectedModule === 'ai' || selectedModule === 'auto_agent'">默认端口: 1234 (LM Studio)</div>
+          </n-space>
+        </n-alert>
+
+        <n-space v-else justify="end" style="margin-top: 16px;">
+          <n-button size="small" @click="checkSingleModule(selectedModule)">
+            重新检测
+          </n-button>
+        </n-space>
+      </n-card>
+    </n-modal>
+  </div>
+</template>
+
+<script setup lang="ts">
+import { ref, computed, onMounted, onUnmounted } from 'vue'
+import { useRouter } from 'vue-router'
+import axios from 'axios'
+import {
+  CheckmarkCircleOutline as CheckmarkCircleIcon,
+  CloseCircleOutline as CloseCircleIcon,
+  RefreshOutline as RefreshIcon,
+  TerminalOutline as TerminalIcon,
+  DownloadOutline as DownloadIcon,
+  CloudUploadOutline as UploadIcon,
+  MedicalOutline as MedicalIcon,
+  WarningOutline as WarningIcon,
+  TimeOutline as TimeIcon,
+  HelpCircleOutline as HelpCircleIcon
+} from '@vicons/ionicons5'
+import { useMessage } from 'naive-ui'
+import CoreControl from '@/components/CoreControl.vue'
+import { useWebSocketStore } from '@/stores/websocket'
+import { useDiagnosticsStore } from '@/stores/diagnostics'
+
+const router = useRouter()
+const message = useMessage()
+const wsStore = useWebSocketStore()
+const diagStore = useDiagnosticsStore()
+
+const checkingAll = ref(false)
+const showDetails = ref(false)
+const showModuleModal = ref(false)
+const selectedModule = ref<string>('')
+
+const systemStats = computed(() => wsStore.systemStats)
+const lastReport = computed(() => diagStore.lastReport)
+const moduleHealthDisplay = computed(() => diagStore.moduleHealth)
+
+const memoryPercent = computed(() => Math.round(systemStats.value?.memory?.percent ?? 0))
+const uptimeFormatted = computed(() => {
+  const seconds = systemStats.value?.process?.uptime ?? 0
+  const hours = Math.floor(seconds / 3600)
+  const mins = Math.floor((seconds % 3600) / 60)
+  return hours > 0 ? `${hours}h ${mins}m` : `${mins}m`
+})
+
+const statusColor = computed(() => {
+  if (diagStore.overallStatus === 'success') return '#18a058'
+  if (diagStore.overallStatus === 'warning') return '#f0a020'
+  return '#d03050'
+})
+
+const statusText = computed(() => {
+  if (diagStore.overallStatus === 'success') return '系统健康'
+  if (diagStore.overallStatus === 'warning') return '部分异常'
+  return '需要关注'
+})
+
+const currentModuleHealth = computed(() => {
+  if (!selectedModule.value) return null
+  return diagStore.moduleHealth[selectedModule.value]
+})
+
+const getProgressColor = (value: number) => {
+  if (value < 60) return '#18a058'
+  if (value < 80) return '#f0a020'
+  return '#d03050'
+}
+
+const getModuleName = (key: string) => {
+  const names: Record<string, string> = {
+    asr: '语音识别 (ASR)',
+    tts: '语音合成 (TTS)',
+    ai: 'AI 对话',
+    auto_agent: '自动代理',
+    llm_core: 'LLM 核心'
+  }
+  return names[key] || key
+}
+
+const getDefaultPort = (module: string) => {
+  const ports: Record<string, string> = {
+    asr: '20260',
+    tts: '20261',
+    ai: '1234',
+    auto_agent: '1234',
+    llm_core: '本地'
+  }
+  return ports[module] || '-'
+}
+
+const formatTime = (timestamp: number) => {
+  if (!timestamp) return '从未'
+  const diff = Date.now() - timestamp
+  if (diff < 60000) return '刚刚'
+  if (diff < 3600000) return `${Math.floor(diff/60000)}分钟前`
+  return new Date(timestamp).toLocaleTimeString()
+}
+
+const runSystemDiagnostics = async () => {
+  try {
+    await diagStore.runFullDiagnostics()
+    message.success('系统体检完成')
+  } catch (error) {
+    message.error('体检失败: ' + (error as Error).message)
+  }
+}
+
+const checkSingleModule = async (name: string) => {
+  await diagStore.checkModule(name)
+}
+
+const refreshAllModules = async () => {
+  checkingAll.value = true
+  await diagStore.checkAllModules()
+  checkingAll.value = false
+  message.success('模块状态已刷新')
+}
+
+const showModuleDetails = (name: string) => {
+  selectedModule.value = name
+  showModuleModal.value = true
+}
+
+const handleReloadConfig = async () => {
+  try {
+    const res = await axios.post('/api/config/reload')
+    if (res.data.success) message.success('配置已重载')
+  } catch {
+    message.error('重载配置失败')
+  }
+}
+
+const exportConfig = () => {
+  window.open('/api/config/export', '_blank')
+}
+
+const handleImportFinish = () => {
+  message.success('配置导入成功，正在重载...')
+  setTimeout(() => window.location.reload(), 1000)
+}
+
+const handleImportError = () => {
+  message.error('配置导入失败')
+}
+
+let heartbeatCleanup: (() => void) | null = null
+
+onMounted(() => {
+  wsStore.connect()
+  diagStore.initWebSocketListeners()
+  heartbeatCleanup = wsStore.startHeartbeat()
+
+  setTimeout(() => {
+    diagStore.checkAllModules()
+    diagStore.startAutoCheck(30000)
+  }, 1000)
+})
+
+onUnmounted(() => {
+  if (heartbeatCleanup) heartbeatCleanup()
+  diagStore.stopAutoCheck()
+})
+</script>
+
+<style scoped>
+.dashboard {
+  max-width: 1400px;
+  margin: 0 auto;
+}
+
+.diagnostics-card {
+  margin-bottom: 16px;
+  background: linear-gradient(135deg, var(--n-card-color) 0%, var(--n-action-color) 100%);
+  border-left: 4px solid v-bind(statusColor);
+}
+
+.stat-card {
+  background: linear-gradient(135deg, var(--n-card-color) 0%, var(--n-card-color) 100%);
+  transition: transform 0.3s;
+}
+
+.stat-card:hover {
+  transform: translateY(-2px);
+}
+
+.stat-detail {
+  margin-top: 8px;
+  font-size: 12px;
+  color: var(--n-text-color-3);
+}
+
+.module-card,
+.action-card {
+  margin-top: 24px;
+}
+
+.module-item {
+  display: flex;
+  align-items: center;
+  gap: 12px;
+  padding: 16px;
+  border-radius: 8px;
+  background-color: var(--n-action-color);
+  transition: all 0.3s;
+  cursor: pointer;
+}
+
+.module-item:hover {
+  background-color: var(--n-hover-color);
+}
+
+.module-item.active {
+  background-color: rgba(24, 160, 88, 0.1);
+  border: 1px solid rgba(24, 160, 88, 0.3);
+}
+
+.module-item.error {
+  background-color: rgba(208, 48, 80, 0.1);
+  border: 1px solid rgba(208, 48, 80, 0.3);
+}
+
+.module-item.checking {
+  background-color: rgba(240, 160, 32, 0.1);
+  border: 1px solid rgba(240, 160, 32, 0.3);
+}
+
+.module-info {
+  flex: 1;
+  min-width: 0;
+}
+
+.module-name {
+  font-weight: 500;
+  font-size: 14px;
+}
+
+.module-status {
+  font-size: 12px;
+  color: var(--n-text-color-3);
+  margin-top: 4px;
+  white-space: nowrap;
+  overflow: hidden;
+  text-overflow: ellipsis;
+}
+
+.module-status.healthy {
+  color: #18a058;
+}
+
+.module-status.unhealthy {
+  color: #d03050;
+}
+
+.module-status.checking {
+  color: #f0a020;
+}
+
+.latency {
+  font-size: 11px;
+  opacity: 0.8;
+  margin-left: 4px;
+  font-family: monospace;
+}
+
+.text-secondary {
+  font-size: 12px;
+  color: var(--n-text-color-3);
+}
+</style>

src/server_view/frontend/src/views/Devices.vue

@@ -0,0 +1,229 @@
+<template>
+  <div class="devices-page">
+    <n-space vertical :size="16">
+      <n-h2>
+        <n-text type="primary">嵌入式设备管理</n-text>
+      </n-h2>
+
+      <n-space justify="space-between" align="center">
+        <n-button @click="refreshDevices" :loading="loading" type="primary">
+          <template #icon>
+            <n-icon><refresh-icon /></n-icon>
+          </template>
+          刷新设备列表
+        </n-button>
+        <n-tag v-if="devices.length > 0" type="success" round>
+          在线设备: {{ devices.length }}
+        </n-tag>
+      </n-space>
+
+      <n-empty v-if="!loading && devices.length === 0" description="暂无在线设备">
+        <template #icon>
+          <n-icon size="48"><device-icon /></n-icon>
+        </template>
+      </n-empty>
+
+      <n-grid v-else :cols="1" :x-gap="12" :y-gap="12">
+        <n-grid-item v-for="device in devices" :key="device.device_id">
+          <n-card :title="device.name" hoverable>
+            <template #header-extra>
+              <n-space>
+                <n-tag type="success" round size="small">在线</n-tag>
+                <n-popconfirm @positive-click="handleRemoveDevice(device.device_id)">
+                  <template #trigger>
+                    <n-button circle size="small" type="error">
+                      <template #icon><n-icon><close-icon /></n-icon></template>
+                    </n-button>
+                  </template>
+                  确定移除设备 {{ device.name }}?
+                </n-popconfirm>
+              </n-space>
+            </template>
+
+            <n-descriptions label-placement="left" bordered :column="1" size="small">
+              <n-descriptions-item label="设备ID">
+                <n-text code>{{ device.device_id }}</n-text>
+              </n-descriptions-item>
+              <n-descriptions-item label="描述">
+                {{ device.description || '无描述' }}
+              </n-descriptions-item>
+              <n-descriptions-item label="注册时间">
+                {{ formatTime(device.register_time) }}
+              </n-descriptions-item>
+            </n-descriptions>
+
+            <n-divider />
+
+            <n-h5>可用功能</n-h5>
+            <n-empty v-if="!device.functions || device.functions.length === 0" description="无可用功能" />
+            <n-space v-else>
+              <n-tag v-for="fn in device.functions" :key="fn.name" type="info" size="small">
+                {{ fn.name }}
+              </n-tag>
+            </n-space>
+
+            <template #action>
+              <n-space justify="center">
+                <n-button @click="showRpcModal(device)" type="primary" size="small">
+                  <template #icon><n-icon><terminal-icon /></n-icon></template>
+                  发送RPC命令
+                </n-button>
+              </n-space>
+            </template>
+          </n-card>
+        </n-grid-item>
+      </n-grid>
+    </n-space>
+
+    <!-- RPC 发送对话框 -->
+    <n-modal v-model:show="rpcModalVisible" title="发送 RPC 命令" preset="card" style="width: 600px">
+      <n-space vertical>
+        <n-alert type="info" :title="'目标设备: ' + (rpcTarget?.name || '')" closable>
+          输入 JSON-RPC 2.0 调用字符串
+        </n-alert>
+        <n-input
+          v-model:value="rpcPayload"
+          type="textarea"
+          :rows="6"
+          placeholder='{"method": "set_speed", "params": {"value": 100}, "id": 1}'
+        />
+        <n-space justify="end">
+          <n-button @click="rpcModalVisible = false">取消</n-button>
+          <n-button @click="handleSendRpc" type="primary" :loading="rpcSending">
+            发送
+          </n-button>
+        </n-space>
+      </n-space>
+    </n-modal>
+  </div>
+</template>
+
+<script setup lang="ts">
+import { ref, onMounted, onUnmounted } from 'vue'
+import { useMessage } from 'naive-ui'
+import { useWebSocketStore } from '@/stores/websocket'
+import {
+  RefreshOutline as RefreshIcon,
+  HardwareChipOutline as DeviceIcon,
+  CloseOutline as CloseIcon,
+  TerminalOutline as TerminalIcon
+} from '@vicons/ionicons5'
+
+const message = useMessage()
+const wsStore = useWebSocketStore()
+
+const loading = ref(false)
+const devices = ref<any[]>([])
+const rpcModalVisible = ref(false)
+const rpcTarget = ref<any>(null)
+const rpcPayload = ref('')
+const rpcSending = ref(false)
+
+function formatTime(t: number | string) {
+  if (!t) return '未知'
+  const d = typeof t === 'number' ? new Date(t * 1000) : new Date(t)
+  return d.toLocaleString()
+}
+
+function refreshDevices() {
+  loading.value = true
+  const socket = wsStore.socket
+  if (socket?.connected) {
+    socket.emit('get_devices')
+  } else {
+    fetchDevicesHttp()
+  }
+}
+
+async function fetchDevicesHttp() {
+  try {
+    const res = await fetch('/api/devices')
+    const json = await res.json()
+    if (json.success) {
+      devices.value = json.data || []
+    }
+  } catch (e) {
+    console.error('获取设备列表失败', e)
+  } finally {
+    loading.value = false
+  }
+}
+
+function showRpcModal(device: any) {
+  rpcTarget.value = device
+  rpcPayload.value = ''
+  rpcModalVisible.value = true
+}
+
+async function handleSendRpc() {
+  if (!rpcTarget.value || !rpcPayload.value) return
+  rpcSending.value = true
+  const socket = wsStore.socket
+  if (socket?.connected) {
+    socket.emit('send_device_rpc', {
+      device_id: rpcTarget.value.device_id,
+      rpc_call: rpcPayload.value
+    })
+  }
+  rpcSending.value = false
+  rpcModalVisible.value = false
+}
+
+function handleRemoveDevice(deviceId: string) {
+  devices.value = devices.value.filter(d => d.device_id !== deviceId)
+}
+
+function onDevicesList(data: any) {
+  loading.value = false
+  if (data.success) {
+    devices.value = data.data || []
+  }
+}
+
+function onDeviceRpcResult(data: any) {
+  if (data.success) {
+    message.success(data.message || 'RPC 命令已发送到设备')
+  } else {
+    message.error(data.error || 'RPC 发送失败')
+  }
+}
+
+function onDeviceRpcResponse(data: any) {
+  const payload = data.payload || {}
+  const result = payload.result
+  const error = payload.error
+  const devId = data.device_id || ''
+
+  if (error) {
+    message.error(`设备 ${devId} RPC 返回错误: ${JSON.stringify(error)}`)
+  } else {
+    message.success(`设备 ${devId} RPC 返回: ${JSON.stringify(result)}`)
+  }
+}
+
+onMounted(() => {
+  const socket = wsStore.socket
+  if (socket) {
+    socket.on('devices_list', onDevicesList)
+    socket.on('device_rpc_result', onDeviceRpcResult)
+    socket.on('device_rpc_response', onDeviceRpcResponse)
+  }
+  refreshDevices()
+})
+
+onUnmounted(() => {
+  const socket = wsStore.socket
+  if (socket) {
+    socket.off('devices_list', onDevicesList)
+    socket.off('device_rpc_result', onDeviceRpcResult)
+    socket.off('device_rpc_response', onDeviceRpcResponse)
+  }
+})
+</script>
+
+<style scoped>
+.devices-page {
+  max-width: 1200px;
+  margin: 0 auto;
+}
+</style>

src/server_view/frontend/src/views/Logs.vue

@@ -0,0 +1,765 @@
+<template>
+  <div class="logs-terminal-page">
+    <!-- 终端工具栏 -->
+    <div class="terminal-toolbar">
+      <n-space align="center" justify="space-between" style="width: 100%">
+        <n-space align="center">
+          <n-icon size="20" color="#18a058"><terminal-icon /></n-icon>
+          <span class="toolbar-title">系统日志终端</span>
+          <n-divider vertical />
+
+          <!-- 日志统计 -->
+          <n-tag size="small" :type="logStats.error > 0 ? 'error' : 'default'">
+            ERROR: {{ logStats.error }}
+          </n-tag>
+          <n-tag size="small" type="warning">
+            WARN: {{ logStats.warn }}
+          </n-tag>
+          <n-tag size="small" type="info">
+            INFO: {{ logStats.info }}
+          </n-tag>
+          <span class="total-logs">总计: {{ wsStore.logs.length }} 行</span>
+        </n-space>
+
+        <n-space align="center">
+          <!-- 自动跟随开关（终端核心功能） -->
+          <n-tooltip>
+            <template #trigger>
+              <n-button
+                :type="isFollowing ? 'primary' : 'default'"
+                size="small"
+                @click="toggleFollow"
+                :class="{ 'following': isFollowing }"
+              >
+                <template #icon>
+                  <n-icon>
+                    <caret-down-icon v-if="isFollowing" />
+                    <pause-icon v-else />
+                  </n-icon>
+                </template>
+                {{ isFollowing ? '跟随' : '暂停' }}
+              </n-button>
+            </template>
+            <span>{{ isFollowing ? '自动滚动到底部' : '已暂停自动滚动' }}</span>
+          </n-tooltip>
+
+          <!-- 等级筛选 -->
+          <n-select
+            v-model:value="selectedLevel"
+            :options="levelOptions"
+            style="width: 100px"
+            size="small"
+            @update:value="handleLevelChange"
+          />
+
+          <!-- 搜索 -->
+          <n-input
+            v-model:value="searchText"
+            placeholder="搜索日志..."
+            size="small"
+            clearable
+            style="width: 150px"
+          >
+            <template #prefix>
+              <n-icon><search-icon /></n-icon>
+            </template>
+          </n-input>
+
+          <n-divider vertical />
+
+          <n-button size="small" @click="clearLogs" type="error" quaternary>
+            <template #icon><n-icon><trash-icon /></n-icon></template>
+            清空
+          </n-button>
+
+          <n-button size="small" @click="downloadLogs">
+            <template #icon><n-icon><download-icon /></n-icon></template>
+            导出
+          </n-button>
+
+          <n-tag :type="wsStore.status === 'connected' ? 'success' : 'error'" size="small">
+            {{ wsStore.status === 'connected' ? '● 实时' : '● 离线' }}
+          </n-tag>
+        </n-space>
+      </n-space>
+    </div>
+
+    <!-- 终端主体区域 -->
+    <div class="terminal-container" ref="terminalContainer">
+      <!-- 新日志提示条（当不在底部且有新日志时显示） -->
+      <div
+        v-show="hasNewLogs && !isFollowing"
+        class="new-logs-indicator"
+        @click="scrollToBottom"
+      >
+        <n-icon><arrow-down-icon /></n-icon>
+        <span>{{ newLogsCount }} 条新日志 - 点击滚动到底部</span>
+      </div>
+
+      <!-- 日志内容区域（终端风格） -->
+      <div
+        class="terminal-content"
+        ref="logContent"
+        @scroll="handleScroll"
+      >
+        <div
+          v-for="(log, index) in filteredLogs"
+          :key="index"
+          class="terminal-line"
+          :class="getLogLevelClass(log)"
+        >
+          <!-- 时间戳 -->
+          <span class="timestamp" v-if="showTimestamp">
+            {{ extractTimestamp(log) }}
+          </span>
+
+          <!-- 日志等级标签 -->
+          <span class="level-badge" :class="getLogLevelClass(log)">
+            {{ getLogLevelLabel(log) }}
+          </span>
+
+          <!-- 日志内容（可点击选择复制） -->
+          <span
+            class="log-message"
+            v-html="highlightSearch(log)"
+            @dblclick="selectLine(log)"
+          ></span>
+        </div>
+
+        <!-- 空状态 -->
+        <div v-if="filteredLogs.length === 0" class="empty-terminal">
+          <n-empty description="暂无日志输出" size="small">
+            <template #icon>
+              <n-icon :size="40" color="#333"><terminal-outline-icon /></n-icon>
+            </template>
+          </n-empty>
+        </div>
+
+        <!-- 底部锚点（用于自动滚动定位） -->
+        <div ref="bottomAnchor" class="bottom-anchor"></div>
+      </div>
+    </div>
+
+    <!-- 终端状态栏 -->
+    <div class="terminal-statusbar">
+      <n-space align="center" justify="space-between" style="width: 100%">
+        <n-space align="center" size="small">
+          <span v-if="selectedLine" class="selected-info">
+            已选择: {{ selectedLine.substring(0, 50) }}{{ selectedLine.length > 50 ? '...' : '' }}
+          </span>
+          <span v-else class="hint-text">
+            提示: 双击行选择文本 | Ctrl+F 搜索 | 滚轮查看历史
+          </span>
+        </n-space>
+
+        <n-space align="center" size="small">
+          <span class="status-item">滚动位置: {{ scrollPercent }}%</span>
+          <span class="status-item">显示: {{ filteredLogs.length }}/{{ wsStore.logs.length }}</span>
+          <span class="status-item" :class="{ 'following-active': isFollowing }">
+            {{ isFollowing ? '● 跟随模式' : '○ 手动浏览' }}
+          </span>
+        </n-space>
+      </n-space>
+    </div>
+
+    <!-- 行详情弹窗（双击行时显示完整内容） -->
+    <n-modal v-model:show="showLineDetail" title="日志详情" style="width: 800px">
+      <n-card>
+        <div class="line-detail-content">{{ selectedLine }}</div>
+        <n-space justify="end" style="margin-top: 16px;">
+          <n-button @click="copySelectedLine">复制</n-button>
+          <n-button @click="showLineDetail = false">关闭</n-button>
+        </n-space>
+      </n-card>
+    </n-modal>
+  </div>
+</template>
+
+<script setup lang="ts">
+import { ref, computed, watch, nextTick, onMounted, onUnmounted } from 'vue'
+import {
+  TrashOutline as TrashIcon,
+  DownloadOutline as DownloadIcon,
+  SearchOutline as SearchIcon,
+  TerminalOutline as TerminalIcon,
+  CaretDownOutline as CaretDownIcon,
+  PauseOutline as PauseIcon,
+  ArrowDownOutline as ArrowDownIcon
+} from '@vicons/ionicons5'
+import { useWebSocketStore } from '@/stores/websocket'
+import { useMessage, useDialog } from 'naive-ui'
+
+const wsStore = useWebSocketStore()
+const message = useMessage()
+const dialog = useDialog()
+
+// DOM引用
+const terminalContainer = ref<HTMLElement>()
+const logContent = ref<HTMLElement>()
+const bottomAnchor = ref<HTMLElement>()
+
+// 状态
+const selectedLevel = ref('ALL')
+const searchText = ref('')
+const isFollowing = ref(true)  // 终端核心：是否自动跟随底部
+const hasNewLogs = ref(false)  // 有新日志但未跟随
+const newLogsCount = ref(0)
+const showTimestamp = ref(true)
+const selectedLine = ref('')
+const showLineDetail = ref(false)
+const scrollPercent = ref(100)
+
+// 用于判断是否在底部的阈值（像素）
+const BOTTOM_THRESHOLD = 50
+
+// 日志等级定义
+const LOG_LEVELS = [
+  { key: 'DEBUG', label: 'DEBUG', color: '#569cd6', weight: 1 },
+  { key: 'INFO', label: 'INFO', color: '#4ec9b0', weight: 2 },
+  { key: 'SUCCESS', label: 'OK', color: '#4ec9b0', weight: 2 },
+  { key: 'WARNING', label: 'WARN', color: '#dcdcaa', weight: 3 },
+  { key: 'ERROR', label: 'ERR', color: '#f44747', weight: 4 },
+  { key: 'CRITICAL', label: 'CRIT', color: '#f44747', weight: 5 }
+]
+
+const levelOptions = [
+  { label: '全部', value: 'ALL' },
+  { label: '调试', value: 'DEBUG' },
+  { label: '信息', value: 'INFO' },
+  { label: '警告', value: 'WARNING' },
+  { label: '错误', value: 'ERROR' },
+  { label: '严重', value: 'CRITICAL' }
+]
+
+// 统计各等级日志数量
+const logStats = computed(() => {
+  const stats = { error: 0, warn: 0, info: 0, debug: 0 }
+  wsStore.logs.forEach(log => {
+    const level = getLogLevel(log)
+    if (level === 'error' || level === 'critical') stats.error++
+    else if (level === 'warning') stats.warn++
+    else if (level === 'info' || level === 'success') stats.info++
+    else stats.debug++
+  })
+  return stats
+})
+
+// 过滤后的日志
+const filteredLogs = computed(() => {
+  let logs = [...wsStore.logs]
+
+  // 等级过滤
+  if (selectedLevel.value !== 'ALL') {
+    const targetLevel = selectedLevel.value.toLowerCase()
+    logs = logs.filter(log => {
+      const level = getLogLevel(log)
+      if (selectedLevel.value === 'ERROR') {
+        return level === 'error' || level === 'critical'
+      }
+      return level === targetLevel
+    })
+  }
+
+  // 搜索过滤
+  if (searchText.value) {
+    const keyword = searchText.value.toLowerCase()
+    logs = logs.filter(log => log.toLowerCase().includes(keyword))
+  }
+
+  return logs
+})
+
+// 获取日志等级
+const getLogLevel = (log: string): string => {
+  for (const level of LOG_LEVELS) {
+    if (log.includes(`| ${level.key} |`) ||
+        log.includes(`[${level.key}]`) ||
+        log.includes(` ${level.key} `)) {
+      return level.key.toLowerCase()
+    }
+  }
+  return 'other'
+}
+
+const getLogLevelClass = (log: string): string => {
+  return getLogLevel(log)
+}
+
+const getLogLevelLabel = (log: string): string => {
+  const level = getLogLevel(log)
+  if (level === 'other') return 'LOG'
+  const found = LOG_LEVELS.find(l => l.key.toLowerCase() === level)
+  return found ? found.label : 'LOG'
+}
+
+// 提取时间戳
+const extractTimestamp = (log: string): string => {
+  const match = log.match(/^(\d{4}-\d{2}-\d{2}\s+\d{2}:\d{2}:\d{2}(?:\.\d+)?)/)
+  return match ? match[1].split(' ')[1] : '' // 只返回时间部分
+}
+
+// 高亮搜索词
+const highlightSearch = (log: string) => {
+  if (!searchText.value) return escapeHtml(log)
+  const regex = new RegExp(`(${escapeRegex(searchText.value)})`, 'gi')
+  return escapeHtml(log).replace(regex, '<mark>$1</mark>')
+}
+
+const escapeHtml = (text: string): string => {
+  const div = document.createElement('div')
+  div.textContent = text
+  return div.innerHTML
+}
+
+const escapeRegex = (string: string): string => {
+  return string.replace(/[.*+?^${}()|[\]\\]/g, '\\$&')
+}
+
+// 核心：滚动到底部（终端行为）
+const scrollToBottom = async (smooth = false) => {
+  await nextTick()
+  if (bottomAnchor.value) {
+    bottomAnchor.value.scrollIntoView({
+      behavior: smooth ? 'smooth' : 'auto',
+      block: 'end'
+    })
+  }
+  hasNewLogs.value = false
+  newLogsCount.value = 0
+  scrollPercent.value = 100
+}
+
+// 检查是否在底部
+const checkIfNearBottom = (): boolean => {
+  if (!logContent.value) return true
+  const { scrollTop, scrollHeight, clientHeight } = logContent.value
+  const distanceToBottom = scrollHeight - scrollTop - clientHeight
+  return distanceToBottom < BOTTOM_THRESHOLD
+}
+
+// 处理滚动事件
+const handleScroll = () => {
+  if (!logContent.value) return
+
+  const { scrollTop, scrollHeight, clientHeight } = logContent.value
+  const distanceToBottom = scrollHeight - scrollTop - clientHeight
+
+  // 计算滚动百分比
+  scrollPercent.value = Math.round(((scrollTop + clientHeight) / scrollHeight) * 100)
+
+  // 如果用户手动滚动离开底部，暂停自动跟随
+  if (distanceToBottom > BOTTOM_THRESHOLD && isFollowing.value) {
+    isFollowing.value = false
+    console.log('用户手动滚动，暂停跟随')
+  }
+
+  // 如果用户滚动到底部附近，恢复跟随
+  if (distanceToBottom < BOTTOM_THRESHOLD && !isFollowing.value) {
+    isFollowing.value = true
+    hasNewLogs.value = false
+    newLogsCount.value = 0
+  }
+}
+
+// 切换跟随模式
+const toggleFollow = () => {
+  isFollowing.value = !isFollowing.value
+  if (isFollowing.value) {
+    scrollToBottom(true)
+    message.success('已恢复自动跟随')
+  } else {
+    message.info('已暂停自动跟随，可自由查看历史日志')
+  }
+}
+
+// 监听日志变化（核心：自动滚动逻辑）
+watch(() => wsStore.logs.length, (newLength, oldLength) => {
+  if (newLength > oldLength) {
+    const added = newLength - oldLength
+    if (!isFollowing.value) {
+      // 不跟随状态下累积新日志计数
+      hasNewLogs.value = true
+      newLogsCount.value += added
+    } else {
+      // 跟随状态下自动滚到底部
+      scrollToBottom(false)
+    }
+  }
+})
+
+// 监听过滤条件变化（重新计算后保持在当前位置或底部）
+watch([selectedLevel, searchText], () => {
+  if (isFollowing.value) {
+    nextTick(() => scrollToBottom(false))
+  }
+})
+
+// 处理等级变更
+const handleLevelChange = (level: string) => {
+  wsStore.subscribeWithLevel(level)
+}
+
+// 清空日志
+const clearLogs = () => {
+  dialog.warning({
+    title: '清空日志',
+    content: '确定要清空所有日志吗？此操作不可恢复。',
+    positiveText: '确定',
+    negativeText: '取消',
+    onPositiveClick: () => {
+      wsStore.clearLogs()
+      hasNewLogs.value = false
+      newLogsCount.value = 0
+      message.success('日志已清空')
+    }
+  })
+}
+
+// 下载日志
+const downloadLogs = () => {
+  const blob = new Blob([wsStore.logs.join('\n')], { type: 'text/plain' })
+  const url = URL.createObjectURL(blob)
+  const a = document.createElement('a')
+  a.href = url
+  a.download = `yosuga-logs-${new Date().toISOString().slice(0,19).replace(/:/g,'-')}.log`
+  document.body.appendChild(a)
+  a.click()
+  document.body.removeChild(a)
+  URL.revokeObjectURL(url)
+}
+
+// 双击选择行
+const selectLine = (log: string) => {
+  selectedLine.value = log
+  showLineDetail.value = true
+}
+
+const copySelectedLine = () => {
+  navigator.clipboard.writeText(selectedLine.value)
+  message.success('已复制到剪贴板')
+}
+
+// 键盘快捷键
+const handleKeydown = (e: KeyboardEvent) => {
+  if (e.ctrlKey && e.key === 'f') {
+    e.preventDefault()
+    // 聚焦搜索框
+    const searchInput = document.querySelector('.logs-terminal-page input') as HTMLInputElement
+    searchInput?.focus()
+  }
+  if (e.key === 'Escape') {
+    isFollowing.value = true
+    scrollToBottom()
+  }
+}
+
+onMounted(() => {
+  // 初始滚动到底部
+  setTimeout(() => scrollToBottom(false), 100)
+
+  // 监听键盘
+  window.addEventListener('keydown', handleKeydown)
+
+  // 定期更新滚动位置显示（用于状态栏）
+  const updateInterval = setInterval(() => {
+    if (logContent.value && isFollowing.value) {
+      scrollPercent.value = 100
+    }
+  }, 1000)
+
+  return () => clearInterval(updateInterval)
+})
+
+onUnmounted(() => {
+  window.removeEventListener('keydown', handleKeydown)
+})
+</script>
+
+<style scoped>
+/* 终端页面布局 */
+.logs-terminal-page {
+  height: calc(100vh - 140px);
+  display: flex;
+  flex-direction: column;
+  background: #1e1e1e; /* VS Code终端深色背景 */
+  border-radius: 8px;
+  overflow: hidden;
+  border: 1px solid #333;
+}
+
+/* 工具栏 */
+.terminal-toolbar {
+  padding: 12px 16px;
+  background: #252526;
+  border-bottom: 1px solid #333;
+  display: flex;
+  align-items: center;
+}
+
+.toolbar-title {
+  font-weight: 600;
+  color: #cccccc;
+  font-size: 14px;
+}
+
+.total-logs {
+  font-size: 12px;
+  color: #858585;
+  margin-left: 8px;
+}
+
+/* 终端主体 */
+.terminal-container {
+  flex: 1;
+  position: relative;
+  overflow: hidden;
+  display: flex;
+  flex-direction: column;
+}
+
+/* 新日志提示条 */
+.new-logs-indicator {
+  position: absolute;
+  top: 0;
+  left: 0;
+  right: 0;
+  background: rgba(24, 160, 88, 0.9);
+  color: white;
+  text-align: center;
+  padding: 8px;
+  cursor: pointer;
+  z-index: 10;
+  display: flex;
+  align-items: center;
+  justify-content: center;
+  gap: 8px;
+  font-size: 13px;
+  font-weight: 500;
+  transition: opacity 0.3s;
+  backdrop-filter: blur(4px);
+}
+
+.new-logs-indicator:hover {
+  background: rgba(24, 160, 88, 1);
+}
+
+/* 日志内容区域 - 终端风格 */
+.terminal-content {
+  flex: 1;
+  overflow-y: auto;
+  overflow-x: hidden;
+  padding: 12px 16px;
+  font-family: 'JetBrains Mono', 'Fira Code', 'Consolas', 'Courier New', monospace;
+  font-size: 13px;
+  line-height: 1.6;
+  background: #1e1e1e;
+  color: #d4d4d4;
+
+  /* 终端滚动条样式 */
+  scrollbar-width: thin;
+  scrollbar-color: #424242 #1e1e1e;
+}
+
+.terminal-content::-webkit-scrollbar {
+  width: 10px;
+  height: 10px;
+}
+
+.terminal-content::-webkit-scrollbar-track {
+  background: #1e1e1e;
+}
+
+.terminal-content::-webkit-scrollbar-thumb {
+  background: #424242;
+  border-radius: 5px;
+}
+
+.terminal-content::-webkit-scrollbar-thumb:hover {
+  background: #4f4f4f;
+}
+
+/* 终端行样式 */
+.terminal-line {
+  display: flex;
+  align-items: flex-start;
+  gap: 12px;
+  padding: 2px 0;
+  white-space: pre-wrap;
+  word-break: break-all;
+  border-radius: 2px;
+  transition: background-color 0.1s;
+}
+
+.terminal-line:hover {
+  background-color: rgba(255, 255, 255, 0.03);
+}
+
+/* 时间戳 */
+.timestamp {
+  color: #858585;
+  font-size: 12px;
+  min-width: 80px;
+  flex-shrink: 0;
+  user-select: none;
+}
+
+/* 等级标签 */
+.level-badge {
+  display: inline-flex;
+  align-items: center;
+  justify-content: center;
+  min-width: 45px;
+  padding: 1px 6px;
+  border-radius: 3px;
+  font-size: 11px;
+  font-weight: 600;
+  flex-shrink: 0;
+  font-family: monospace;
+  text-transform: uppercase;
+}
+
+.level-badge.debug {
+  background-color: rgba(86, 156, 214, 0.15);
+  color: #569cd6;
+}
+
+.level-badge.info,
+.level-badge.success {
+  background-color: rgba(78, 201, 176, 0.15);
+  color: #4ec9b0;
+}
+
+.level-badge.warning {
+  background-color: rgba(220, 220, 170, 0.15);
+  color: #dcdcaa;
+}
+
+.level-badge.error,
+.level-badge.critical {
+  background-color: rgba(244, 71, 71, 0.15);
+  color: #f44747;
+}
+
+.level-badge.other {
+  background-color: rgba(133, 133, 133, 0.15);
+  color: #858585;
+}
+
+/* 日志消息 */
+.log-message {
+  flex: 1;
+  color: #d4d4d4;
+}
+
+/* 错误行高亮 */
+.terminal-line.error .log-message,
+.terminal-line.critical .log-message {
+  color: #f44747;
+}
+
+.terminal-line.warning .log-message {
+  color: #dcdcaa;
+}
+
+.terminal-line.debug .log-message {
+  color: #858585;
+}
+
+/* 搜索高亮 */
+:deep(mark) {
+  background-color: rgba(220, 220, 170, 0.4);
+  color: #dcdcaa;
+  padding: 0 2px;
+  border-radius: 2px;
+  font-weight: bold;
+}
+
+/* 空状态 */
+.empty-terminal {
+  display: flex;
+  align-items: center;
+  justify-content: center;
+  height: 100%;
+  color: #858585;
+}
+
+/* 底部锚点 */
+.bottom-anchor {
+  height: 1px;
+  flex-shrink: 0;
+}
+
+/* 状态栏 */
+.terminal-statusbar {
+  padding: 8px 16px;
+  background: #252526;
+  border-top: 1px solid #333;
+  font-size: 12px;
+  color: #858585;
+  display: flex;
+  align-items: center;
+}
+
+.status-item {
+  padding: 0 8px;
+  border-right: 1px solid #333;
+}
+
+.status-item:last-child {
+  border-right: none;
+}
+
+.following-active {
+  color: #4ec9b0;
+}
+
+.hint-text {
+  color: #6e6e6e;
+  font-style: italic;
+}
+
+.selected-info {
+  color: #4ec9b0;
+  max-width: 400px;
+  overflow: hidden;
+  text-overflow: ellipsis;
+  white-space: nowrap;
+}
+
+/* 跟随按钮高亮 */
+.following {
+  animation: pulse 2s infinite;
+}
+
+@keyframes pulse {
+  0%, 100% { opacity: 1; }
+  50% { opacity: 0.7; }
+}
+
+/* 行详情弹窗 */
+.line-detail-content {
+  background: #1e1e1e;
+  color: #d4d4d4;
+  padding: 16px;
+  border-radius: 4px;
+  font-family: 'JetBrains Mono', monospace;
+  font-size: 13px;
+  line-height: 1.6;
+  white-space: pre-wrap;
+  word-break: break-all;
+  max-height: 400px;
+  overflow: auto;
+  border: 1px solid #333;
+}
+
+/* 响应式 */
+@media (max-width: 768px) {
+  .timestamp {
+    display: none;
+  }
+
+  .terminal-toolbar .n-space:first-child {
+    display: none;
+  }
+}
+</style>

src/server_view/frontend/src/views/Monitor.vue

@@ -0,0 +1,197 @@
+<template>
+  <div class="monitor-page">
+    <n-grid :x-gap="16" :y-gap="16" cols="1 l:2">
+      <n-gi>
+        <n-card title="CPU 使用率历史" :bordered="false">
+          <v-chart class="chart" :option="cpuChartOption" autoresize />
+        </n-card>
+      </n-gi>
+
+      <n-gi>
+        <n-card title="内存使用历史" :bordered="false">
+          <v-chart class="chart" :option="memoryChartOption" autoresize />
+        </n-card>
+      </n-gi>
+    </n-grid>
+
+    <n-card title="实时指标" class="metrics-card" :bordered="false">
+      <n-descriptions bordered :column="3">
+        <n-descriptions-item label="CPU 核心数">
+          {{ systemStats.cpu.count }}
+        </n-descriptions-item>
+        <n-descriptions-item label="内存总量">
+          {{ formatBytes(systemStats.memory.total) }}
+        </n-descriptions-item>
+        <n-descriptions-item label="磁盘使用率">
+          {{ systemStats.disk.percent?.toFixed(1) }}%
+        </n-descriptions-item>
+        <n-descriptions-item label="进程内存占用">
+          {{ systemStats.process.memory_percent?.toFixed(2) }}%
+        </n-descriptions-item>
+        <n-descriptions-item label="进程 CPU 占用">
+          {{ systemStats.process.cpu_percent?.toFixed(1) }}%
+        </n-descriptions-item>
+        <n-descriptions-item label="线程数">
+          {{ systemStats.process.threads }}
+        </n-descriptions-item>
+      </n-descriptions>
+    </n-card>
+  </div>
+</template>
+
+<script setup lang="ts">
+import { ref, computed, onMounted, onUnmounted } from 'vue'
+import axios from 'axios'
+import { use } from 'echarts/core'
+import { CanvasRenderer } from 'echarts/renderers'
+import { LineChart } from 'echarts/charts'
+import {
+  GridComponent,
+  TooltipComponent,
+  LegendComponent,
+  TitleComponent
+} from 'echarts/components'
+import VChart from 'vue-echarts'
+
+use([
+  CanvasRenderer,
+  LineChart,
+  GridComponent,
+  TooltipComponent,
+  LegendComponent,
+  TitleComponent
+])
+
+const systemStats = ref({
+  cpu: { percent: 0, count: 0 },
+  memory: { total: 0, percent: 0 },
+  disk: { percent: 0 },
+  process: { memory_percent: 0, cpu_percent: 0, threads: 0 }
+})
+
+const cpuHistory = ref<number[]>([])
+const memoryHistory = ref<number[]>([])
+const timeLabels = ref<string[]>([])
+
+const cpuChartOption = computed(() => ({
+  tooltip: { trigger: 'axis' },
+  grid: { left: '3%', right: '4%', bottom: '3%', containLabel: true },
+  xAxis: {
+    type: 'category',
+    boundaryGap: false,
+    data: timeLabels.value
+  },
+  yAxis: {
+    type: 'value',
+    max: 100,
+    axisLabel: { formatter: '{value}%' }
+  },
+  series: [{
+    name: 'CPU',
+    type: 'line',
+    smooth: true,
+    areaStyle: {
+      color: {
+        type: 'linear',
+        x: 0, y: 0, x2: 0, y2: 1,
+        colorStops: [
+          { offset: 0, color: 'rgba(24, 160, 88, 0.3)' },
+          { offset: 1, color: 'rgba(24, 160, 88, 0.05)' }
+        ]
+      }
+    },
+    lineStyle: { color: '#18a058' },
+    itemStyle: { color: '#18a058' },
+    data: cpuHistory.value
+  }]
+}))
+
+const memoryChartOption = computed(() => ({
+  tooltip: { trigger: 'axis' },
+  grid: { left: '3%', right: '4%', bottom: '3%', containLabel: true },
+  xAxis: {
+    type: 'category',
+    boundaryGap: false,
+    data: timeLabels.value
+  },
+  yAxis: {
+    type: 'value',
+    max: 100,
+    axisLabel: { formatter: '{value}%' }
+  },
+  series: [{
+    name: '内存',
+    type: 'line',
+    smooth: true,
+    areaStyle: {
+      color: {
+        type: 'linear',
+        x: 0, y: 0, x2: 0, y2: 1,
+        colorStops: [
+          { offset: 0, color: 'rgba(32, 128, 240, 0.3)' },
+          { offset: 1, color: 'rgba(32, 128, 240, 0.05)' }
+        ]
+      }
+    },
+    lineStyle: { color: '#2080f0' },
+    itemStyle: { color: '#2080f0' },
+    data: memoryHistory.value
+  }]
+}))
+
+const formatBytes = (bytes: number) => {
+  if (bytes === 0) return '0 B'
+  const k = 1024
+  const sizes = ['B', 'KB', 'MB', 'GB', 'TB']
+  const i = Math.floor(Math.log(bytes) / Math.log(k))
+  return parseFloat((bytes / Math.pow(k, i)).toFixed(2)) + ' ' + sizes[i]
+}
+
+let intervalId: NodeJS.Timeout
+
+const fetchStats = async () => {
+  try {
+    const res = await axios.get('/api/system/info')
+    if (res.data.success) {
+      systemStats.value = res.data.data
+
+      cpuHistory.value.push(res.data.data.cpu.percent)
+      memoryHistory.value.push(res.data.data.memory.percent)
+      timeLabels.value.push(new Date().toLocaleTimeString())
+
+      if (cpuHistory.value.length > 20) {
+        cpuHistory.value.shift()
+        memoryHistory.value.shift()
+        timeLabels.value.shift()
+      }
+    }
+  } catch (error) {
+    console.error('获取监控数据失败:', error)
+  }
+}
+
+onMounted(() => {
+  fetchStats()
+  intervalId = setInterval(fetchStats, 2000)
+})
+
+onUnmounted(() => {
+  clearInterval(intervalId)
+})
+</script>
+
+<style scoped>
+.monitor-page {
+  max-width: 1400px;
+  margin: 0 auto;
+}
+
+.chart {
+  height: 300px;
+  width: 100%;
+}
+
+.metrics-card {
+  margin-top: 24px;
+}
+</style>

src/server_view/frontend/tsconfig.json

@@ -0,0 +1,25 @@
+{
+  "compilerOptions": {
+    "target": "ES2020",
+    "useDefineForClassFields": true,
+    "module": "ESNext",
+    "lib": ["ES2020", "DOM", "DOM.Iterable"],
+    "skipLibCheck": true,
+    "moduleResolution": "bundler",
+    "allowImportingTsExtensions": true,
+    "resolveJsonModule": true,
+    "isolatedModules": true,
+    "noEmit": true,
+    "jsx": "preserve",
+    "strict": true,
+    "noUnusedLocals": true,
+    "noUnusedParameters": true,
+    "noFallthroughCasesInSwitch": true,
+    "baseUrl": ".",
+    "paths": {
+      "@/*": ["./src/*"]
+    }
+  },
+  "include": ["src/**/*.ts", "src/**/*.tsx", "src/**/*.vue"],
+  "references": [{ "path": "./tsconfig.node.json" }]
+}

src/server_view/frontend/tsconfig.node.json

@@ -0,0 +1,11 @@
+{
+  "compilerOptions": {
+    "composite": true,
+    "skipLibCheck": true,
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "allowSyntheticDefaultImports": true,
+    "strict": true
+  },
+  "include": ["vite.config.ts"]
+}

src/server_view/frontend/vite.config.ts

@@ -0,0 +1,31 @@
+import { defineConfig } from 'vite'
+import vue from '@vitejs/plugin-vue'
+import { resolve } from 'path'
+
+export default defineConfig({
+  plugins: [vue()],
+  resolve: {
+    alias: {
+      '@': resolve(__dirname, 'src'),
+    },
+  },
+  server: {
+    port: 5173,
+    proxy: {
+      '/api': {
+        target: 'http://localhost:8089',
+        changeOrigin: true,
+      },
+      '/socket.io': {
+        target: 'http://localhost:8089',
+        changeOrigin: true,
+        ws: true,
+      },
+    },
+  },
+  build: {
+    outDir: '../backend/static',
+    emptyOutDir: true,
+    assetsDir: 'assets',
+  },
+})

src/server_view/start_web.py

@@ -0,0 +1,14 @@
+"""
+Yosuga Server Web UI 启动脚本
+"""
+import sys
+from pathlib import Path
+
+# 添加项目根目录到路径
+project_root = Path(__file__).parent.parent.parent
+sys.path.insert(0, str(project_root))
+
+from backend.app import run_server
+
+if __name__ == '__main__':
+    run_server(host="0.0.0.0", port=8089, debug=False)