1. 增加了跨平台的截屏与系统信息获取工具模块

2. 增加了对于服务端截屏请求的处理类

3rdparty/autogui-cpp/src/Autogui.cpp

@@ -11,6 +11,10 @@
 #include <random>
 #include <thread>
 
+#ifndef M_PI
+# define M_PI		3.14159265358979323846
+#endif
+
 namespace AutoGUI {
 
 // 内部辅助函数

src/Core/Inc/AppCore.h

@@ -9,3 +9,31 @@
  *  1. 处理来自服务端的数据，分发并执行
  *  2. 完成非阻塞的事件循环处理，构建业务状态机
  */
+#include <QMutex>
+#include <QObject>
+
+class AppCore final : public QObject {
+Q_OBJECT
+Q_DISABLE_COPY(AppCore)     // 禁用拷贝
+
+private:
+    /**
+     * 构造函数私有化
+     * @param parent
+     */
+    explicit AppCore(QObject *parent = nullptr);        // 并不将本模块挂在对象树当中，因为本模块为单例类，内存自行管理
+
+    static QScopedPointer<AppCore> m_instance;          // 单例类
+    static QMutex m_mutex;
+private slots:
+    // 业务接收槽函数
+
+public:
+    // 单例访问点
+    static AppCore *getInstance();
+    // 显式销毁
+    static void destroy();
+
+    ~AppCore() override;
+
+};

src/Core/Inc/GLCore.h

@@ -3,6 +3,9 @@
 #include <QtWidgets/QWidget>
 #include <QOpenGLWidget>
 #include "menu.h"
+#ifdef Q_OS_WIN
+#include <windows.h>
+#endif
 
 class GLCore final : public QOpenGLWidget
 {
@@ -81,7 +84,6 @@ private:
     bool isRightPressed;            /// 鼠标右键是否按下
     QPoint currentPos;              /// 当前鼠标位置
 #ifdef Q_OS_WIN
-#include <windows.h>
 private:
     HWND hwnd;  // Windows窗口句柄
     void setWindowTransparentForMouse(bool transparent);

src/Core/Src/AppCore.cpp

@@ -3,3 +3,52 @@
 //
 
 #include "AppCore.h"
+#include <QDebug>
+
+#include "AudioDataHandle.h"
+#include "AutoAgentHandle.h"
+#include "ScreenShotReqDataHandle.h"
+
+// 初始化静态成员
+QScopedPointer<AppCore> AppCore::m_instance;
+QMutex AppCore::m_mutex;
+
+// 单例实现 (QScopedPointer + Mutex)
+AppCore* AppCore::getInstance()
+{
+    if (m_instance.isNull()) {
+        QMutexLocker locker(&m_mutex);
+        if (m_instance.isNull()) {
+            // 使用 reset 创建实例，因为构造函数是私有的
+            m_instance.reset(new AppCore());
+        }
+    }
+    return m_instance.data();
+}
+
+void AppCore::destroy()
+{
+    QMutexLocker locker(&m_mutex);
+    if (!m_instance.isNull()) {
+        m_instance.reset(); // 这会触发析构函数
+    }
+}
+
+AppCore::AppCore(QObject *parent) : QObject(parent)
+{
+    // 初始化业务解析单例
+    AudioDataHandle::getInstance();
+    AutoAgentHandle::getInstance();
+    ScreenShotReqDataHandle::getInstance();
+}
+
+AppCore::~AppCore()
+{
+    // 析构业务解析单例
+    ScreenShotReqDataHandle::destroy();
+    AutoAgentHandle::destroy();     // 显式销毁
+    AudioDataHandle::destroy();
+
+
+    qDebug() << "AppCore destroyed";
+}

src/DAO/Inc/NetWorkDO.h

@@ -28,6 +28,7 @@
 #include "DataTransferObjectBase.h"
 #include "AudioDataTransferObject.h"
 #include "AutoAgentDataObject.h"
+#include "ScreenShotDataTransferObject.h"
 /**
  * NetworkDO
  */
@@ -54,8 +55,9 @@ public:
 
 signals:
     // 业务接收信号
-    void audioPacketReceived(const AudioDataTransferObject& packet);    // 音频数据准备完成信号
+    void audioPacketReceived(const AudioDataTransferObject& packet);            // 音频数据准备完成信号
-    void autoAgentPacketReceived(const AutoAgentDataObject& packet);    // 自动代理数据包接收信号
+    void autoAgentPacketReceived(const AutoAgentDataObject& packet);            // 自动代理数据包接收信号
+    void screenShotPacketReceived(const ScreenShotDataTransferObject& packet);  // 截图数据包接收信号
 
     void errorOccurred(const QString& errorMsg);                        // 错误信号
 

src/DAO/Inc/ScreenShotDataTransferObject.h

@@ -0,0 +1,55 @@
+//
+// Created by misaki on 2026/2/1.
+//
+
+/**
+ * 数据传输对象 (DTO) 定义
+ * ScreenShotDataTransferObject
+ * 与Yosuga_server中的是对等DTO
+ */
+
+#pragma once
+#include <QObject>
+#include <QByteArray>
+#include <QJsonObject>
+#include <QJsonValue>
+#include "DataTransferObjectBase.h"
+// 前向声明，减少依赖
+class QJsonObject;
+class ScreenShotDataTransferObject final : public DataTransferObjectBase{
+public:
+    // 构造函数（带默认值）
+    explicit ScreenShotDataTransferObject(QString  owner = "client",
+                                            bool isSuccess = true,
+                                            QString  realtimeScreenShot = "",
+                                            int width = 0, int height = 0,
+                                            QString  describeInfo = "", QString LLMResponse =  ""
+                                           );
+    // 静态工厂方法
+    static ScreenShotDataTransferObject fromJson(const QJsonObject& json);
+
+    [[nodiscard]] QString type() const override { return "screenshot_req"; }
+
+    // 序列化
+    [[nodiscard]] QJsonObject toJson() const override;    // 通过多态即可统一调用方式
+
+    // 链式调用设置
+    ScreenShotDataTransferObject& setData(const QString& key, const QJsonValue& value) override;
+
+    [[nodiscard]] QString owner() const { return m_owner; }
+    [[nodiscard]] bool isSuccess() const { return m_isSuccess; }
+    [[nodiscard]] QString realtimeScreenShot() const { return m_realtimeScreenShot; }
+    [[nodiscard]] int width() const { return m_width; }
+    [[nodiscard]] int height() const { return m_height; }
+    [[nodiscard]] QString describeInfo() const { return m_describeInfo; }
+    [[nodiscard]] QString LLMResponse() const { return m_LLMResponse; }
+
+private:
+    QString m_owner;                        /// 数据的拥有者(server or client)
+    bool    m_isSuccess;                    /// 截图是否成功
+    QString m_realtimeScreenShot;           /// 客户端设备的实时截图数据(base64)
+    int     m_width;                        /// 截图宽度 非必要字段
+    int     m_height;                       /// 截图高度 非必要字段
+    QString m_describeInfo;                 /// 设备的描述信息(告知模型以做出更加准确的判断) 非必要字段
+    QString m_LLMResponse;                  /// LLM的响应结果(由服务端发送时携带)
+};

src/DAO/Src/NetWorkDO.cpp

@@ -65,13 +65,16 @@ void NetworkDO::onDataReceived(const QString& type, const QJsonObject& data)
     // 根据类型分发数据包
     // 为什么分发做在这里，而不是统一数据再去分发，如果不在这里做分发通知，分开发信号，而使用统一的信号
     // 如果有多个观察者，让观察者自动识别数据包，这会导致信号广播，容易引起性能问题(因为这里依赖的是Qt的信号与槽机制)
-    // TODO: 在此处使用工厂模式，根据type内容快速创建对应的对象
+    // TODO: 考虑在此处使用工厂模式，根据type内容快速创建对应的对象
     if (type == "audio_data") {
         emit audioPacketReceived(AudioDataTransferObject::fromJson(data));      // 构造并发送音频对象
     }
     else if (type == "auto_agent") {
         emit autoAgentPacketReceived(AutoAgentDataObject::fromJson(data));
     }
+    else if (type == "screenshot_req") {
+        emit screenShotPacketReceived(ScreenShotDataTransferObject::fromJson(data));
+    }
     else {
         qWarning() << "[NetworkDO] Received unknown type:" << type;
     }

src/DAO/Src/ScreenShotDataTransferObject.cpp

@@ -0,0 +1,75 @@
+//
+// Created by misaki on 2026/2/1.
+//
+#include "ScreenShotDataTransferObject.h"
+#include <QJsonValue>
+#include <utility>
+#include <QDebug>
+
+// 构造函数实现（初始化列表）
+ScreenShotDataTransferObject::ScreenShotDataTransferObject(QString owner,
+                                                         bool isSuccess,
+                                                         QString realtimeScreenShot,
+                                                         int width, int height,
+                                                         QString describeInfo, QString LLMResponse)
+    : m_owner(std::move(owner))
+    , m_isSuccess(isSuccess)
+    , m_realtimeScreenShot(std::move(realtimeScreenShot))
+    , m_width(width)
+    , m_height(height)
+    , m_describeInfo(std::move(describeInfo))
+    , m_LLMResponse(std::move(LLMResponse))
+{}
+
+// 静态工厂方法：从 JSON 反序列化
+ScreenShotDataTransferObject ScreenShotDataTransferObject::fromJson(const QJsonObject& json) {
+    // 逐个字段读取，不存在则用默认值
+    const QString owner = json.value("Owner").toString("client");
+    const bool isSuccess = json.value("isSuccess").toBool(false);
+    const QString realtimeScreenShot = json.value("RealTimeScreenShot").toString();
+    const int width = json.value("Width").toInt(0);
+    const int height = json.value("Height").toInt(0);
+    const QString describeInfo = json.value("DescribeInfo").toString();
+    const QString LLMResponse = json.value("LLMResponse").toString();
+
+    // 调用构造函数创建对象
+    return ScreenShotDataTransferObject(owner, isSuccess, realtimeScreenShot,
+                                      width, height, describeInfo);
+}
+
+// 序列化为 JSON
+QJsonObject ScreenShotDataTransferObject::toJson() const {
+    QJsonObject json;
+    json["Owner"] = m_owner;
+    json["isSuccess"] = m_isSuccess;
+    json["RealTimeScreenShot"] = m_realtimeScreenShot;
+    json["Width"] = m_width;
+    json["Height"] = m_height;
+    json["DescribeInfo"] = m_describeInfo;
+    json["LLMResponse"] = m_LLMResponse;
+    return json;
+}
+
+// 链式设置
+ScreenShotDataTransferObject& ScreenShotDataTransferObject::setData(const QString& key,
+                                                                    const QJsonValue& value) {
+    if (key == "Owner") {
+        m_owner = value.toString();
+    } else if (key == "isSuccess") {
+        m_isSuccess = value.toBool();
+    } else if (key == "RealTimeScreenShot") {
+        m_realtimeScreenShot = value.toString();
+    } else if (key == "Width") {
+        m_width = value.toInt();
+    } else if (key == "Height") {
+        m_height = value.toInt();
+    } else if (key == "DescribeInfo") {
+        m_describeInfo = value.toString();
+    } else if (key == "LLMResponse") {
+        m_LLMResponse = value.toString();
+    } else {
+        qWarning() << "Unknown key:" << key << "for ScreenShotDataTransferObject";
+    }
+
+    return *this;  // 返回自身引用，支持链式调用
+}

src/Handle/AudioHandle/Src/AudioInput.cpp

@@ -53,7 +53,7 @@ void AudioInput::setAudioSettings(const int rate, const int channels)
 {
     m_format.setSampleRate(rate);
     m_format.setChannelCount(channels);
-    // 重要：Qt6 默认可能是 float，为了生成标准 WAV 且方便计算 RMS，强制设为 Int16
+    // 为了生成标准 WAV 且方便计算 RMS，强制设为 Int16
     m_format.setSampleFormat(QAudioFormat::Int16);
 
     // 检查设备是否支持该格式，不支持则使用最接近的

src/Handle/DataObjectHandle/Inc/AudioDataHandle.h

@@ -13,7 +13,7 @@ class AudioDataHandle final : public QObject
 {
 Q_OBJECT
 Q_DISABLE_COPY(AudioDataHandle)     // 禁用拷贝
-    private:
+private:
     /**
      * 构造函数私有化
      * @param parent

src/Handle/DataObjectHandle/Inc/ScreenShotReqDataHandle.h

@@ -0,0 +1,38 @@
+//
+// Created by misaki on 2026/2/1.
+//
+#pragma once
+#include <QObject>
+#include <QMutex>
+#include "ScreenShotDataTransferObject.h"
+class ScreenShotReqDataHandle final : public QObject
+{
+Q_OBJECT
+Q_DISABLE_COPY(ScreenShotReqDataHandle)     // 禁用拷贝
+    private:
+    /**
+     * 构造函数私有化
+     * @param parent
+     */
+    explicit ScreenShotReqDataHandle(QObject *parent = nullptr);        // 并不将本模块挂在对象树当中，因为本模块为单例类，内存自行管理
+
+    static QScopedPointer<ScreenShotReqDataHandle> m_instance;       // 单例类
+    static QMutex m_mutex;
+
+private slots:
+    // 业务接收槽函数，当获取到截图数据包时，进行解析并处理
+    void onScreenShotPacketReceived(const ScreenShotDataTransferObject& packet) const;
+
+    signals:
+    // 发送截图处理完成的信号，供界面显示使用
+    void screenShotProcessed(const QPixmap& screenshot, const QString& description);
+public:
+    // 单例访问点
+    static ScreenShotReqDataHandle *getInstance();
+    // 显式销毁
+    static void destroy();
+
+    ~ScreenShotReqDataHandle() override;
+private:
+    QString m_systemInfo;
+};

src/Handle/DataObjectHandle/Src/AudioDataHandle.cpp

@@ -42,15 +42,15 @@ AudioDataHandle::~AudioDataHandle()
 
 void AudioDataHandle::onAudioPacketReceived(const AudioDataTransferObject &packet) {
     // 管理并调用AudioOutput播放流式wav音频
-    if (packet.isEnd()) {   // 如果是结束包
+    if (packet.isEnd()) {   // 如果是结束包(空包)
         AudioOutput::getInstance()->stopStream();   // 停止播放
         return;
     }
-    if (packet.isStart()) { // 如果是开始包
+    if (packet.isStart()) { // 如果是开始包(单wav 44字节头)
         AudioOutput::getInstance()->startStream(packet.sampleRate(), packet.channelCount(), packet.bitDepth());;  // 播放开始
         return;
     }
-    // 否则播放即可
+    // 否则加入播放队列即可
     AudioOutput::getInstance()->pushStreamData(packet.audioData());
 }
 

src/Handle/DataObjectHandle/Src/ScreenShotReqDataHandle.cpp

@@ -0,0 +1,66 @@
+//
+// Created by misaki on 2026/2/1.
+//
+#include "ScreenShotReqDataHandle.h"
+#include "NetWorkDO.h"
+#include <QDebug>
+#include <QPixmap>
+#include "ScreenShotDataTransferObject.h"
+#include "ScreenHelperUtil.hpp"
+// 初始化静态成员
+QScopedPointer<ScreenShotReqDataHandle> ScreenShotReqDataHandle::m_instance;
+QMutex ScreenShotReqDataHandle::m_mutex;
+
+// 单例实现 (QScopedPointer + Mutex)
+ScreenShotReqDataHandle* ScreenShotReqDataHandle::getInstance()
+{
+    if (m_instance.isNull()) {
+        QMutexLocker locker(&m_mutex);
+        if (m_instance.isNull()) {
+            // 使用 reset 创建实例，因为构造函数是私有的
+            m_instance.reset(new ScreenShotReqDataHandle());
+        }
+    }
+    return m_instance.data();
+}
+
+void ScreenShotReqDataHandle::destroy()
+{
+    QMutexLocker locker(&m_mutex);
+    if (!m_instance.isNull()) {
+        m_instance.reset(); // 这会触发析构函数
+    }
+}
+
+
+ScreenShotReqDataHandle::ScreenShotReqDataHandle(QObject *parent) : QObject(parent)
+{
+    connect(NetworkDO::getInstance(), &NetworkDO::screenShotPacketReceived,
+            this, &ScreenShotReqDataHandle::onScreenShotPacketReceived);
+    // 初始化时候就构造好关于当前运行平台的信息
+    ScreenHelper::SystemInfo sysInfo = ScreenHelper::getSystemInfo();
+    const QString sysText = QString("System: %1  OS Version: %2  Display Server: %3")
+                          .arg(sysInfo.osType, sysInfo.osVersion, sysInfo.displayServer);
+    this->m_systemInfo = sysText;
+}
+
+ScreenShotReqDataHandle::~ScreenShotReqDataHandle()
+{
+    qDebug() << "ScreenShotDataHandle destroyed";
+}
+
+void ScreenShotReqDataHandle::onScreenShotPacketReceived(const ScreenShotDataTransferObject &packet) const {
+    qDebug() << "ScreenShot packet request from:" << packet.owner();
+    // 截图当前画面并构造对等DTO发送
+    const ScreenHelper::ScreenshotResult result = ScreenHelper::captureFocusedScreen();   // 获取当前屏幕截图
+    if (!result.success) {  // 如果截图失败
+        // TODO: 考虑失败时候构造一个错误DTO给服务端
+        return;
+    }
+    ScreenShotDataTransferObject reback;    // 构造返回的DTO
+    reback.setData("isSuccess", true).setData("RealTimeScreenShot", result.base64Data)
+        .setData("Width", result.width).setData("Height", result.height)
+        .setData("DescribeInfo", this->m_systemInfo);
+    // 发送DTO
+    NetworkDO::getInstance()->sendPacket(reback);
+}

src/UI/Setting/Src/NetworkPage.cpp

@@ -10,7 +10,6 @@
 #include "websocketmanager.h"
 #include "NetWorkDO.h"
 #include <QFile>
-#include "AudioDataHandle.h"
 
 NetWorkPage::NetWorkPage(QWidget* parent)
         : BasePage(parent)
@@ -73,7 +72,6 @@ void NetWorkPage::initUI() {
 void NetWorkPage::initWebSocketClient() {
     auto* client = WebSocketClient::getInstance();     // 获取单例实例(设置一个默认地址)
     auto* netDO = NetworkDO::getInstance();
-    AudioDataHandle::getInstance();         // 初始化音频处理模块
     // 注入：将底层发送能力赋予 NetworkDO
     netDO->registerSender([client](const QString& type, const QJsonObject& data){
         client->sendJson(type, data);

src/Utils/Inc/ScreenHelperUtil.hpp

@@ -0,0 +1,53 @@
+//
+// Created by misaki on 2026/2/1.
+//
+/**
+ * 屏幕截图与系统信息获取工具类
+ */
+
+#pragma once
+#include <QString>
+#include <QSize>
+#include <QScreen>
+#include <QPixmap>
+#include <QBuffer>
+#include <QGuiApplication>
+#include <QWindow>
+#include <QCursor>
+#include <QSysInfo>
+class ScreenHelper
+{
+public:
+    // 系统信息struct
+    struct SystemInfo {
+        QString osType;         // 例如: "windows", "linux", "macos"
+        QString osVersion;      // 例如: "Windows 11 (10.0)", "Ubuntu 22.04"
+        QString displayServer;  // 例如: "windows", "cocoa", "xcb" (X11), "wayland"
+        bool isWayland;         // 专门标记是否为 Wayland
+    };
+
+    // 截图结果struct
+    struct ScreenshotResult {
+        bool success;           // 是否成功
+        QString base64Data;     // 图片的Base64字符串 (PNG格式)
+        int width;              // 图片宽度
+        int height;             // 图片高度
+        QString screenName;     // 屏幕名称
+        QString errorMsg;       // 如果失败，返回错误信息
+    };
+public:
+    /**
+     * @brief 获取当前焦点屏幕的全屏截图并转换为Base64 \n
+     * 判定逻辑：优先取有焦点的窗口所在屏幕，若无，取鼠标所在屏幕
+     */
+    static ScreenshotResult captureFocusedScreen();
+
+    /**
+     * @brief 获取当前操作系统和显示服务信息
+     */
+    static SystemInfo getSystemInfo();
+
+private:
+    // 私有构造，禁止实例化
+    ScreenHelper() = default;
+};

src/Utils/Src/ScreenHelperUtil.cpp

@@ -0,0 +1,92 @@
+//
+// Created by misaki on 2026/2/1.
+//
+#include <QDebug>
+#include "ScreenHelperUtil.hpp"
+ScreenHelper::ScreenshotResult ScreenHelper::captureFocusedScreen()
+{
+    ScreenHelper::ScreenshotResult result;
+    result.success = false;
+    // 获取目标屏幕
+    QScreen *targetScreen = nullptr;
+
+    // 首先尝试获取当前应用程序拥有焦点的窗口所在的屏幕
+    QWindow *focusWindow = QGuiApplication::focusWindow();
+    if (focusWindow) {
+        targetScreen = focusWindow->screen();
+    }
+    // 如果没有窗口焦点或者窗口还没显示，获取鼠标光标所在的屏幕
+    if (!targetScreen) {
+        targetScreen = QGuiApplication::screenAt(QCursor::pos());
+    }
+
+    // 如果以上都失败，回退到主屏幕
+    if (!targetScreen) {
+        targetScreen = QGuiApplication::primaryScreen();
+    }
+
+    if (!targetScreen) {
+        result.errorMsg = "Critical Error: No detectible screen found.";
+        return result;
+    }
+
+    // 获取屏幕基本信息
+    result.screenName = targetScreen->name();
+
+    // 执行截图
+    // grabWindow(0) 表示截取整个屏幕
+    // 注：在 Wayland 上，这可能需要系统权限或会弹出确认框，或者在某些安全策略下返回黑色图像
+    QPixmap pixmap = targetScreen->grabWindow(0);
+
+    if (pixmap.isNull()) {
+        result.errorMsg = "Failed to grab screen content (Permission denied or System restriction).";
+        return result;
+    }
+
+    result.width = pixmap.width();
+    result.height = pixmap.height();
+
+    // 转换为 Base64
+    QByteArray byteArray;
+    QBuffer buffer(&byteArray);
+    buffer.open(QIODevice::WriteOnly);
+    // 保存为 PNG 格式，质量默认即可
+    if (pixmap.save(&buffer, "PNG")) {
+        result.base64Data = QString::fromLatin1(byteArray.toBase64());
+        result.success = true;
+    } else {
+        result.errorMsg = "Failed to encode image to PNG buffer.";
+    }
+
+    return result;
+}
+
+ScreenHelper::SystemInfo ScreenHelper::getSystemInfo()
+{
+    ScreenHelper::SystemInfo info;
+    // 获取操作系统类型
+    info.osType = QSysInfo::productType();
+
+    // 获取详细版本 (例如 Windows 10/11, Ubuntu 20.04)
+    // prettyProductName() 通常能区分 Win10 和 Win11
+    info.osVersion = QSysInfo::prettyProductName();
+
+    // 获取显示服务器类型 (Platform Plugin)
+    // 这里的返回值通常是 QPA 插件的名字
+    // Windows -> "windows"
+    // macOS -> "cocoa"
+    // Linux X11 -> "xcb"
+    // Linux Wayland -> "wayland"
+    QString platformName = QGuiApplication::platformName();
+    info.displayServer = platformName;
+
+    // 专门判断 Wayland
+    info.isWayland = (platformName == "wayland");
+    // 针对 Linux 做更细致的显示名称优化
+    if (platformName == "xcb") {
+        info.displayServer = "X11 (xcb)";
+    } else if (platformName == "wayland") {
+        info.displayServer = "Wayland";
+    }
+    return info;
+}