Add MCP server

Author: 78

main/CMakeLists.txt

@@ -15,6 +15,7 @@ set(SOURCES "audio_codecs/audio_codec.cc"
             "protocols/websocket_protocol.cc"
             "iot/thing.cc"
             "iot/thing_manager.cc"
+            "mcp_server.cc"
             "system_info.cc"
             "application.cc"
             "ota.cc"

main/Kconfig.projbuild

@@ -8,7 +8,7 @@ config OTA_URL
 
 
 choice
-    prompt "语言选择"
+    prompt "Default Language"
     default LANGUAGE_ZH_CN
     help
         Select device display language
@@ -249,37 +249,48 @@ choice DISPLAY_ESP32S3_KORVO2_V3
 endchoice
 
 config USE_WECHAT_MESSAGE_STYLE
-    bool "使用微信聊天界面风格"
+    bool "Enable WeChat Message Style"
     default n
     help
         使用微信聊天界面风格
 
 config USE_WAKE_WORD_DETECT
-    bool "启用唤醒词检测"
+    bool "Enable Wake Word Detection"
     default y
     depends on IDF_TARGET_ESP32S3 || IDF_TARGET_ESP32P4 && SPIRAM
     help
         需要 ESP32 S3 与 AFE 支持
 
 config USE_AUDIO_PROCESSOR
-    bool "启用音频降噪、增益处理"
+    bool "Enable Audio Noise Reduction"
     default y
     depends on IDF_TARGET_ESP32S3 || IDF_TARGET_ESP32P4 && SPIRAM
     help
         需要 ESP32 S3 与 AFE 支持
 
 config USE_DEVICE_AEC
-    bool "在通话过程中启用设备端 AEC"
+    bool "Enable Device-Side AEC"
     default n
     depends on USE_AUDIO_PROCESSOR && (BOARD_TYPE_ESP_BOX_3 || BOARD_TYPE_ESP_BOX || BOARD_TYPE_ESP_BOX_LITE || BOARD_TYPE_LICHUANG_DEV || BOARD_TYPE_ESP32S3_KORVO2_V3 || BOARD_TYPE_ESP32S3_Touch_AMOLED_1_75)
     help
         因为性能不够，不建议和微信聊天界面风格同时开启
 
 config USE_SERVER_AEC
-    bool "在通话过程中启用服务器端 AEC"
+    bool "Enable Server-Side AEC"
     default n
     depends on USE_AUDIO_PROCESSOR
     help
         启用服务器端 AEC，需要服务器支持
 
+choice IOT_PROTOCOL
+    prompt "IoT Protocol"
+    default IOT_PROTOCOL_XIAOZHI
+    help
+        IoT 协议，用于获取设备状态与发送控制指令
+    config IOT_PROTOCOL_MCP
+        bool "MCP协议 2024-11-05"
+    config IOT_PROTOCOL_XIAOZHI
+        bool "小智IoT协议 1.0"
+endchoice
+
 endmenu

main/application.cc

@@ -9,6 +9,7 @@
 #include "font_awesome_symbols.h"
 #include "iot/thing_manager.h"
 #include "assets/lang_config.h"
+#include "mcp_server.h"
 
 #if CONFIG_USE_AUDIO_PROCESSOR
 #include "afe_audio_processor.h"
@@ -41,7 +42,7 @@ static const char* const STATE_STRINGS[] = {
 
 Application::Application() {
     event_group_ = xEventGroupCreate();
-    background_task_ = new BackgroundTask(4096 * 8);
+    background_task_ = new BackgroundTask(4096 * 7);
 
 #if CONFIG_USE_AUDIO_PROCESSOR
     audio_processor_ = std::make_unique<AfeAudioProcessor>();
@@ -425,12 +426,15 @@ void Application::Start() {
                 protocol_->server_sample_rate(), codec->output_sample_rate());
         }
         SetDecodeSampleRate(protocol_->server_sample_rate(), protocol_->server_frame_duration());
+
+#if CONFIG_IOT_PROTOCOL_XIAOZHI
         auto& thing_manager = iot::ThingManager::GetInstance();
         protocol_->SendIotDescriptors(thing_manager.GetDescriptorsJson());
         std::string states;
         if (thing_manager.GetStatesJson(states, false)) {
             protocol_->SendIotStates(states);
         }
+#endif
     });
     protocol_->OnAudioChannelClosed([this, &board]() {
         board.SetPowerSaveMode(true);
@@ -465,7 +469,7 @@ void Application::Start() {
                 });
             } else if (strcmp(state->valuestring, "sentence_start") == 0) {
                 auto text = cJSON_GetObjectItem(root, "text");
-                if (text != NULL) {
+                if (cJSON_IsString(text)) {
                     ESP_LOGI(TAG, "<< %s", text->valuestring);
                     Schedule([this, display, message = std::string(text->valuestring)]() {
                         display->SetChatMessage("assistant", message.c_str());
@@ -474,31 +478,40 @@ void Application::Start() {
             }
         } else if (strcmp(type->valuestring, "stt") == 0) {
             auto text = cJSON_GetObjectItem(root, "text");
-            if (text != NULL) {
+            if (cJSON_IsString(text)) {
                 ESP_LOGI(TAG, ">> %s", text->valuestring);
                 Schedule([this, display, message = std::string(text->valuestring)]() {
                     display->SetChatMessage("user", message.c_str());
                 });
             }
         } else if (strcmp(type->valuestring, "llm") == 0) {
             auto emotion = cJSON_GetObjectItem(root, "emotion");
-            if (emotion != NULL) {
+            if (cJSON_IsString(emotion)) {
                 Schedule([this, display, emotion_str = std::string(emotion->valuestring)]() {
                     display->SetEmotion(emotion_str.c_str());
                 });
             }
+#if CONFIG_IOT_PROTOCOL_MCP
+        } else if (strcmp(type->valuestring, "mcp") == 0) {
+            auto payload = cJSON_GetObjectItem(root, "payload");
+            if (cJSON_IsObject(payload)) {
+                McpServer::GetInstance().ParseMessage(payload);
+            }
+#endif
+#if CONFIG_IOT_PROTOCOL_XIAOZHI
         } else if (strcmp(type->valuestring, "iot") == 0) {
             auto commands = cJSON_GetObjectItem(root, "commands");
-            if (commands != NULL) {
+            if (cJSON_IsArray(commands)) {
                 auto& thing_manager = iot::ThingManager::GetInstance();
                 for (int i = 0; i < cJSON_GetArraySize(commands); ++i) {
                     auto command = cJSON_GetArrayItem(commands, i);
                     thing_manager.Invoke(command);
                 }
             }
+#endif
         } else if (strcmp(type->valuestring, "system") == 0) {
             auto command = cJSON_GetObjectItem(root, "command");
-            if (command != NULL) {
+            if (cJSON_IsString(command)) {
                 ESP_LOGI(TAG, "System command: %s", command->valuestring);
                 if (strcmp(command->valuestring, "reboot") == 0) {
                     // Do a reboot if user requests a OTA update
@@ -513,7 +526,7 @@ void Application::Start() {
             auto status = cJSON_GetObjectItem(root, "status");
             auto message = cJSON_GetObjectItem(root, "message");
             auto emotion = cJSON_GetObjectItem(root, "emotion");
-            if (status != NULL && message != NULL && emotion != NULL) {
+            if (cJSON_IsString(status) && cJSON_IsString(message) && cJSON_IsString(emotion)) {
                 Alert(status->valuestring, message->valuestring, emotion->valuestring, Lang::Sounds::P3_VIBRATION);
             } else {
                 ESP_LOGW(TAG, "Alert command requires status, message and emotion");
@@ -620,7 +633,10 @@ void Application::OnClockTimer() {
     clock_ticks_++;
 
     // Print the debug info every 10 seconds
-    if (clock_ticks_ % 10 == 0) {
+    if (clock_ticks_ % 3 == 0) {
+        // char buffer[500];
+        // vTaskList(buffer);
+        // ESP_LOGI(TAG, "Task list: \n%s", buffer);
         // SystemInfo::PrintRealTimeStats(pdMS_TO_TICKS(1000));
 
         int free_sram = heap_caps_get_free_size(MALLOC_CAP_INTERNAL);
@@ -850,7 +866,9 @@ void Application::SetDeviceState(DeviceState state) {
             display->SetStatus(Lang::Strings::LISTENING);
             display->SetEmotion("neutral");
             // Update the IoT states before sending the start listening command
+#if CONFIG_IOT_PROTOCOL_XIAOZHI
             UpdateIotStates();
+#endif
 
             // Make sure the audio processor is running
             if (!audio_processor_->IsRunning()) {
@@ -910,11 +928,13 @@ void Application::SetDecodeSampleRate(int sample_rate, int frame_duration) {
 }
 
 void Application::UpdateIotStates() {
+#if CONFIG_IOT_PROTOCOL_XIAOZHI
     auto& thing_manager = iot::ThingManager::GetInstance();
     std::string states;
     if (thing_manager.GetStatesJson(states, true)) {
         protocol_->SendIotStates(states);
     }
+#endif
 }
 
 void Application::Reboot() {
@@ -955,3 +975,11 @@ bool Application::CanEnterSleepMode() {
     // Now it is safe to enter sleep mode
     return true;
 }
+
+void Application::SendMcpMessage(const std::string& payload) {
+    Schedule([this, payload]() {
+        if (protocol_) {
+            protocol_->SendMcpMessage(payload);
+        }
+    });
+}

main/application.h

@@ -72,6 +72,7 @@ class Application {
     void WakeWordInvoke(const std::string& wake_word);
     void PlaySound(const std::string_view& sound);
     bool CanEnterSleepMode();
+    void SendMcpMessage(const std::string& payload);
 
 private:
     Application();

main/boards/bread-compact-wifi/compact_wifi_board.cc

@@ -94,6 +94,7 @@ class CompactWifiBoard : public WifiBoard {
             display_ = new NoDisplay();
             return;
         }
+        ESP_ERROR_CHECK(esp_lcd_panel_invert_color(panel_, false));
 
         // Set the display to on
         ESP_LOGI(TAG, "Turning display on");

main/boards/common/board.h

@@ -49,6 +49,7 @@ class Board {
     virtual std::string GetJson();
     virtual void SetPowerSaveMode(bool enabled) = 0;
     virtual std::string GetBoardJson() = 0;
+    virtual std::string GetDeviceStatusJson() = 0;
 };
 
 #define DECLARE_BOARD(BOARD_CLASS_NAME) \

main/boards/common/dual_network_board.cc

@@ -98,4 +98,8 @@ void DualNetworkBoard::SetPowerSaveMode(bool enabled) {
 
 std::string DualNetworkBoard::GetBoardJson() {   
     return current_board_->GetBoardJson();
-} 
+}
+
+std::string DualNetworkBoard::GetDeviceStatusJson() {
+    return current_board_->GetDeviceStatusJson();
+}

main/boards/common/dual_network_board.h

@@ -56,7 +56,7 @@ class DualNetworkBoard : public Board {
     virtual const char* GetNetworkStateIcon() override;
     virtual void SetPowerSaveMode(bool enabled) override;
     virtual std::string GetBoardJson() override;
-    
+    virtual std::string GetDeviceStatusJson() override;
 };
 
 #endif // DUAL_NETWORK_BOARD_H 

main/boards/common/ml307_board.cc

@@ -120,3 +120,86 @@ std::string Ml307Board::GetBoardJson() {
 void Ml307Board::SetPowerSaveMode(bool enabled) {
     // TODO: Implement power save mode for ML307
 }
+
+std::string Ml307Board::GetDeviceStatusJson() {
+    /*
+     * 返回设备状态JSON
+     * 
+     * 返回的JSON结构如下：
+     * {
+     *     "audio_speaker": {
+     *         "volume": 70
+     *     },
+     *     "screen": {
+     *         "brightness": 100,
+     *         "theme": "light"
+     *     },
+     *     "battery": {
+     *         "level": 50,
+     *         "charging": true
+     *     },
+     *     "network": {
+     *         "type": "cellular",
+     *         "carrier": "CHINA MOBILE",
+     *         "csq": 10
+     *     }
+     * }
+     */
+    auto& board = Board::GetInstance();
+    auto root = cJSON_CreateObject();
+
+    // Audio speaker
+    auto audio_speaker = cJSON_CreateObject();
+    auto audio_codec = board.GetAudioCodec();
+    if (audio_codec) {
+        cJSON_AddNumberToObject(audio_speaker, "volume", audio_codec->output_volume());
+    }
+    cJSON_AddItemToObject(root, "audio_speaker", audio_speaker);
+
+    // Screen brightness
+    auto backlight = board.GetBacklight();
+    auto screen = cJSON_CreateObject();
+    if (backlight) {
+        cJSON_AddNumberToObject(screen, "brightness", backlight->brightness());
+    }
+    auto display = board.GetDisplay();
+    if (display && display->height() > 64) { // For LCD display only
+        cJSON_AddStringToObject(screen, "theme", display->GetTheme().c_str());
+    }
+    cJSON_AddItemToObject(root, "screen", screen);
+
+    // Battery
+    int battery_level = 0;
+    bool charging = false;
+    bool discharging = false;
+    if (board.GetBatteryLevel(battery_level, charging, discharging)) {
+        cJSON* battery = cJSON_CreateObject();
+        cJSON_AddNumberToObject(battery, "level", battery_level);
+        cJSON_AddBoolToObject(battery, "charging", charging);
+        cJSON_AddItemToObject(root, "battery", battery);
+    }
+
+    // Network
+    auto network = cJSON_CreateObject();
+    cJSON_AddStringToObject(network, "type", "cellular");
+    cJSON_AddStringToObject(network, "carrier", modem_.GetCarrierName().c_str());
+    int csq = modem_.GetCsq();
+    if (csq == -1) {
+        cJSON_AddStringToObject(network, "signal", "unknown");
+    } else if (csq >= 0 && csq <= 14) {
+        cJSON_AddStringToObject(network, "signal", "very weak");
+    } else if (csq >= 15 && csq <= 19) {
+        cJSON_AddStringToObject(network, "signal", "weak");
+    } else if (csq >= 20 && csq <= 24) {
+        cJSON_AddStringToObject(network, "signal", "medium");
+    } else if (csq >= 25 && csq <= 31) {
+        cJSON_AddStringToObject(network, "signal", "strong");
+    }
+    cJSON_AddItemToObject(root, "network", network);
+
+    auto json_str = cJSON_PrintUnformatted(root);
+    std::string json(json_str);
+    cJSON_free(json_str);
+    cJSON_Delete(root);
+    return json;
+}

main/boards/common/ml307_board.h

@@ -21,6 +21,7 @@ class Ml307Board : public Board {
     virtual const char* GetNetworkStateIcon() override;
     virtual void SetPowerSaveMode(bool enabled) override;
     virtual AudioCodec* GetAudioCodec() override { return nullptr; }
+    virtual std::string GetDeviceStatusJson() override;
 };
 
 #endif // ML307_BOARD_H