chore: v3.4.18

fix: 修复主动概率回复关闭后仍然回复的问题 #317
fix: 尝试修复 gewechat 群聊收不到 at 的回复 #294
2025-02-03 16:14:27 +08:00 · 2025-02-03 14:33:53 +08:00 · 2025-02-03 14:28:14 +08:00 · 2025-02-03 14:17:45 +08:00 · 2025-02-03 14:11:50 +08:00 · 2025-02-03 13:27:56 +08:00
59 changed files with 1237 additions and 492 deletions
@@ -12,6 +12,8 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
    && apt-get clean \
    && rm -rf /var/lib/apt/lists/*

+RUN ln -sf /usr/share/zoneinfo/Asia/Shanghai /etc/localtime
+
 RUN python -m pip install -r requirements.txt

 EXPOSE 6185 
@@ -1,14 +1,12 @@
-
 <p align="center">
+    
+![logo](https://github.com/user-attachments/assets/07649e07-3b8e-4feb-9aa9-bf13af4f3476)

-<img src="https://github.com/user-attachments/assets/de10f24d-cd64-433a-90b8-16c0a60de24a" width=500>

 </p>

 <div align="center">

-<h1>AstrBot</h1>
-
 _✨ 易上手的多平台 LLM 聊天机器人及开发框架 ✨_

 [![GitHub release (latest by date)](https://img.shields.io/github/v/release/Soulter/AstrBot)](https://github.com/Soulter/AstrBot/releases/latest)
@@ -16,9 +14,8 @@ _✨ 易上手的多平台 LLM 聊天机器人及开发框架 ✨_
 <a href="https://hub.docker.com/r/soulter/astrbot"><img alt="Docker pull" src="https://img.shields.io/docker/pulls/soulter/astrbot.svg"/></a>
 <img alt="Static Badge" src="https://img.shields.io/badge/QQ群-322154837-purple">
 [![wakatime](https://wakatime.com/badge/user/915e5316-99c6-4563-a483-ef186cf000c9/project/018e705a-a1a7-409a-a849-3013485e6c8e.svg)](https://wakatime.com/badge/user/915e5316-99c6-4563-a483-ef186cf000c9/project/018e705a-a1a7-409a-a849-3013485e6c8e)
+![Dynamic JSON Badge](https://img.shields.io/badge/dynamic/json?url=https%3A%2F%2Fapi.soulter.top%2Fastrbot%2Fstats&query=v&label=7%E6%97%A5%E6%B6%88%E6%81%AF%E4%B8%8A%E8%A1%8C%E9%87%8F&cacheSeconds=3600)
 [![codecov](https://codecov.io/gh/Soulter/AstrBot/graph/badge.svg?token=FF3P5967B8)](https://codecov.io/gh/Soulter/AstrBot)
-[<img src="https://api.gitsponsors.com/api/badge/img?id=575865240" height="20">](https://api.gitsponsors.com/api/badge/link?p=XEpbdGxlitw/RbcwiTX93UMzNK/jgDYC8NiSzamIPMoKvG2lBFmyXhSS/b0hFoWlBBMX2L5X5CxTDsUdyvcIEHTOfnkXz47UNOZvMwyt5CzbYpq0SEzsSV1OJF1cCo90qC/ZyYKYOWedal3MhZ3ikw==)
-</a>

 <a href="https://astrbot.app/">查看文档</a> ｜
 <a href="https://github.com/Soulter/AstrBot/issues">问题提交</a>
@@ -38,7 +35,7 @@ AstrBot 是一个松耦合、异步、支持多消息平台部署、具有易用
 > [!TIP]
 > 管理面板在线体验 Demo: [https://demo.astrbot.app/](https://demo.astrbot.app/)
 > 
-> 用户名: `astrbot`, 密码: `astrbot`。此 Demo 未配置 LLM，因此无法在聊天页使用大模型。
+> 用户名: `astrbot`, 密码: `astrbot`。未配置 LLM，无法在聊天页使用大模型。（不要再修改 demo 的登录密码了 😭）

 ## ✨ 使用方式

@@ -67,19 +64,31 @@ AstrBot 是一个松耦合、异步、支持多消息平台部署、具有易用

 ## ⚡ 消息平台支持情况

-
 | 平台    | 支持性 | 详情 | 消息类型 |
 | -------- | ------- | ------- | ------ |
-| QQ      | ✔    | 私聊、群聊 | 文字、图片、语音 |
-| QQ 官方API | ✔    | 私聊、群聊，QQ 频道私聊、群聊 | 文字、图片 |
-| 微信    | ✔    | [Gewechat](https://github.com/Devo919/Gewechat)。微信个人号私聊、群聊 | 文字 |
+| QQ(官方机器人接口) | ✔    | 私聊、群聊，QQ 频道私聊、群聊 | 文字、图片 |
+| QQ(OneBot)      | ✔    | 私聊、群聊 | 文字、图片、语音 |
+| 微信(个人号)    | ✔    | 微信个人号私聊、群聊 | 文字、图片、语音 |
 | [Telegram](https://github.com/Soulter/astrbot_plugin_telegram)   | ✔    | 私聊、群聊 | 文字、图片 |
+| 微信(企业微信)    | 🚧    | 计划内 | - |
 | 微信对话开放平台 | 🚧    | 计划内 | - |
 | 飞书   | 🚧    | 计划内 | - |
 | Discord   | 🚧    | 计划内 | - |
 | WhatsApp   | 🚧    | 计划内 | - |
 | 小爱音响   | 🚧    | 计划内 | - |

+# 🦌 接下来的路线图
+
+> [!TIP]
+> 欢迎在 Issue 提出更多建议 <3
+
+- [ ] 完善并保证目前所有平台适配器的功能一致性
+- [ ] 优化插件接口
+- [ ] 默认支持更多 TTS 服务，如 GPT-Sovits
+- [ ] 完善“聊天增强”部分，支持持久化记忆
+- [ ] 规划 i18n
+
+
 ## ❤️ 贡献

 欢迎任何 Issues/Pull Requests！只需要将你的更改提交到此项目 ：)
@@ -131,8 +140,21 @@ _✨ 内置 Web Chat，在线与机器人交互 ✨_
 > [!TIP] 
 > 如果本项目对您的生活 / 工作产生了帮助，或者您关注本项目的未来发展，请给项目 Star，这是我维护这个开源项目的动力 <3

+<div align="center">
+    
 [![Star History Chart](https://api.star-history.com/svg?repos=soulter/astrbot&type=Date)](https://star-history.com/#soulter/astrbot&Date)

+</div>
+
+## Sponsors
+
+[<img src="https://api.gitsponsors.com/api/badge/img?id=575865240" height="20">](https://api.gitsponsors.com/api/badge/link?p=XEpbdGxlitw/RbcwiTX93UMzNK/jgDYC8NiSzamIPMoKvG2lBFmyXhSS/b0hFoWlBBMX2L5X5CxTDsUdyvcIEHTOfnkXz47UNOZvMwyt5CzbYpq0SEzsSV1OJF1cCo90qC/ZyYKYOWedal3MhZ3ikw==)
+
+## Disclaimer
+
+1. The project is protected under the `AGPL-v3` opensource license.
+2. The deployment of WeChat (personal account) utilizes [Gewechat](https://github.com/Devo919/Gewechat) service. AstrBot only guarantees connectivity with Gewechat and recommends using a WeChat account that is not frequently used. In the event of account risk control, the author of this project shall not bear any responsibility.
+3. Please ensure compliance with local laws and regulations when using this project.

 <!-- ## ✨ ATRI [Beta 测试]

@@ -144,5 +166,6 @@ _✨ 内置 Web Chat，在线与机器人交互 ✨_
 4. TTS
    -->

-_アトリは、高性能ですから!_
+
+_私は、高性能ですから!_

@@ -6,6 +6,7 @@ from astrbot.core.star.register import (
    register_platform_adapter_type as platform_adapter_type,
    register_permission_type as permission_type,
    register_on_llm_request as on_llm_request,
+    register_on_llm_response as on_llm_response,
    register_llm_tool as llm_tool,
    register_on_decorating_result as on_decorating_result,
    register_after_message_sent as after_message_sent
@@ -31,5 +32,6 @@ __all__ = [
    'on_llm_request',
    'llm_tool',
    'on_decorating_result',
-    'after_message_sent'
+    'after_message_sent',
+    'on_llm_response'
 ]
@@ -2,4 +2,5 @@ from astrbot.core.platform import (
    AstrMessageEvent, Platform, AstrBotMessage, MessageMember, MessageType, PlatformMetadata
 )

-from astrbot.core.platform.register import register_platform_adapter
+from astrbot.core.platform.register import register_platform_adapter
+from astrbot.core.message.components import *
@@ -1,2 +1,2 @@
 from astrbot.core.provider import Provider, STTProvider, Personality
-from astrbot.core.provider.entites import ProviderRequest, ProviderType, ProviderMetaData
+from astrbot.core.provider.entites import ProviderRequest, ProviderType, ProviderMetaData, LLMResponse
@@ -2,7 +2,7 @@ import os
 import json
 import logging
 import enum
-from .default import DEFAULT_CONFIG
+from .default import DEFAULT_CONFIG, DEFAULT_VALUE_MAP
 from typing import Dict

 ASTRBOT_CONFIG_PATH = "data/cmd_config.json"
@@ -13,29 +13,72 @@ class RateLimitStrategy(enum.Enum):
    DISCARD = "discard"

 class AstrBotConfig(dict):
-    '''从配置文件中加载的配置，支持直接通过点号操作符访问配置项'''
+    '''从配置文件中加载的配置，支持直接通过点号操作符访问根配置项。
    
-    def __init__(self):
+    - 初始化时会将传入的 default_config 与配置文件进行比对，如果配置文件中缺少配置项则会自动插入默认值并进行一次写入操作。会递归检查配置项。
+    - 如果配置文件路径对应的文件不存在，则会自动创建并写入默认配置。
+    - 如果传入了 schema，将会通过 schema 解析出 default_config，此时传入的 default_config 会被忽略。
+    '''
+    
+    def __init__(
+        self, 
+        config_path: str = ASTRBOT_CONFIG_PATH, 
+        default_config: dict = DEFAULT_CONFIG,
+        schema: dict = None
+    ):
        super().__init__()
+        
+        # 调用父类的 __setattr__ 方法，防止保存配置时将此属性写入配置文件
+        object.__setattr__(self, 'config_path', config_path)
+        object.__setattr__(self, 'default_config', default_config)
+        object.__setattr__(self, 'schema', schema)
+        
+        if schema:
+            default_config = self._config_schema_to_default_config(schema)
+        
        if not self.check_exist():
            '''不存在时载入默认配置'''
-            with open(ASTRBOT_CONFIG_PATH, "w", encoding="utf-8-sig") as f:
-                json.dump(DEFAULT_CONFIG, f, indent=4, ensure_ascii=False)
+            with open(config_path, "w", encoding="utf-8-sig") as f:
+                json.dump(default_config, f, indent=4, ensure_ascii=False)

-        with open(ASTRBOT_CONFIG_PATH, "r", encoding="utf-8-sig") as f:
+        with open(config_path, "r", encoding="utf-8-sig") as f:
            conf_str = f.read()
            if conf_str.startswith(u'/ufeff'): # remove BOM
                conf_str = conf_str.encode('utf8')[3:].decode('utf8')
            conf = json.loads(conf_str)
        
        # 检查配置完整性，并插入
-        has_new = self.check_config_integrity(DEFAULT_CONFIG, conf)
+        has_new = self.check_config_integrity(default_config, conf)
        self.update(conf)
        if has_new:
            self.save_config()
            
        self.update(conf)
-            
+
+    def _config_schema_to_default_config(self, schema: dict) -> dict:
+        '''将 Schema 转换成 Config'''
+        conf = {}
+        
+        def _parse_schema(schema: dict, conf: dict):
+            for k, v in schema.items():
+                if v['type'] not in DEFAULT_VALUE_MAP:
+                    raise TypeError(f"不受支持的配置类型 {v['type']}。支持的类型有：{DEFAULT_VALUE_MAP.keys()}")
+                if 'default' in v:
+                    default = v['default']
+                else:
+                    default = DEFAULT_VALUE_MAP[v['type']]
+                
+                if v['type'] == 'object':
+                    conf[k] = {}
+                    _parse_schema(v['items'], conf[k])
+                else:
+                    conf[k] = default
+                    
+        _parse_schema(schema, conf)
+
+        return conf
+
+
    def check_config_integrity(self, refer_conf: Dict, conf: Dict, path=""):
        '''检查配置完整性，如果有新的配置项则返回 True'''
        has_new = False
@@ -61,7 +104,7 @@ class AstrBotConfig(dict):
        '''
        if replace_config:
            self.update(replace_config)
-        with open(ASTRBOT_CONFIG_PATH, "w", encoding="utf-8-sig") as f:
+        with open(self.config_path, "w", encoding="utf-8-sig") as f:
            json.dump(self, f, indent=2, ensure_ascii=False)
            
    def __getattr__(self, item):
@@ -81,4 +124,4 @@ class AstrBotConfig(dict):
        self[key] = value

    def check_exist(self) -> bool:
-        return os.path.exists(ASTRBOT_CONFIG_PATH)
+        return os.path.exists(self.config_path)
@@ -2,7 +2,7 @@
 如需修改配置，请在 `data/cmd_config.json` 中修改或者在管理面板中可视化修改。
 """

-VERSION = "3.4.12"
+VERSION = "3.4.18"
 DB_PATH = "data/data_v3.db"

 # 默认配置
@@ -24,13 +24,20 @@ DEFAULT_CONFIG = {
        "wl_ignore_admin_on_friend": True,
        "reply_with_mention": False,
        "reply_with_quote": False,
-        "path_mapping": []
+        "path_mapping": [],
+        "segmented_reply": {
+            "enable": False,
+            "only_llm_result": True,
+            "interval": "1.5,3.5",
+            "regex": ".*?[。？！~…]+|.+$"
+        }
    },
    "provider": [],
    "provider_settings": {
        "enable": True,
        "wake_prefix": "",
        "web_search": False,
+        "web_search_link": False,
        "identifier": False,
        "datetime_system_prompt": True,
        "default_personality": "default",
@@ -44,6 +51,19 @@ DEFAULT_CONFIG = {
        "enable": False,
        "provider_id": "",
    },
+    "provider_ltm_settings": {
+        "group_icl_enable": False,
+        "group_message_max_cnt": 300,
+        "image_caption": False,
+        "image_caption_prompt": "Please describe the image using Chinese.",
+        "active_reply": {
+            "enable": False,
+            "method": "possibility_reply",
+            "possibility_reply": 0.1,
+            "prompt": "",
+        },
+        "put_history_to_prompt": True,
+    },
    "content_safety": {
        "internal_keywords": {"enable": True, "extra_keywords": []},
        "baidu_aip": {"enable": False, "app_id": "", "api_key": "", "secret_key": ""},
@@ -63,14 +83,7 @@ DEFAULT_CONFIG = {
    "pip_install_arg": "",
    "plugin_repo_mirror": "",
    "knowledge_db": {},
-    "persona": [
-        {
-            "name": "default",
-            "prompt": "如果用户寻求帮助或者打招呼，请告诉他可以用 /help 查看 AstrBot 帮助。",
-            "begin_dialogs": [],
-            "mood_imitation_dialogs": [],
-        }
-    ],
+    "persona": [],
 }


@@ -99,14 +112,13 @@ CONFIG_METADATA_2 = {
                        "ws_reverse_host": "",
                        "ws_reverse_port": 6199,
                    },
-                    "vchat(微信)": {"id": "default", "type": "vchat", "enable": False},
                    "gewechat(微信)": {
                        "id": "gwchat",
                        "type": "gewechat",
                        "enable": False,
                        "base_url": "http://localhost:2531",
                        "nickname": "soulter",
-                        "host": "localhost",
+                        "host": "这里填写你的局域网IP或者公网服务器IP",
                        "port": 11451,
                    },
                },
@@ -182,6 +194,31 @@ CONFIG_METADATA_2 = {
                            },
                        },
                    },
+                    "segmented_reply": {
+                        "description": "分段回复",
+                        "type": "object",
+                        "items": {
+                            "enable": {
+                                "description": "启用分段回复",
+                                "type": "bool",
+                            },
+                            "only_llm_result": {
+                                "description": "仅对 LLM 结果分段",
+                                "type": "bool",
+                            },
+                            "interval": {
+                                "description": "随机间隔时间(秒)",
+                                "type": "string",
+                                "hint": "每一段回复的间隔时间，格式为 `最小时间,最大时间`。如 `0.75,2.5`",
+                            },
+                            "regex": {
+                                "description": "正则表达式",
+                                "type": "string",
+                                "obvious_hint": True,
+                                "hint": "用于分隔一段消息。默认情况下会根据句号、问号等标点符号分隔。re.findall(r'<regex>', text)",
+                            },
+                        },
+                    },
                    "reply_prefix": {
                        "description": "回复前缀",
                        "type": "string",
@@ -199,8 +236,9 @@ CONFIG_METADATA_2 = {
                    "id_whitelist": {
                        "description": "ID 白名单",
                        "type": "list",
-                        "items": {"type": "int"},
-                        "hint": "填写后，将只处理所填写的 ID 发来的消息事件。为空时表示不启用白名单过滤。可以使用 /myid 指令获取在某个平台上的会话 ID。也可在 AstrBot 日志内获取会话 ID，当一条消息没通过白名单时，会输出 INFO 级别的日志。会话 ID 类似 aiocqhttp:GroupMessage:547540978",
+                        "items": {"type": "string"},
+                        "obvious_hint": True,
+                        "hint": "AstrBot 只处理所填写的 ID 发来的消息事件。为空时不启用白名单过滤。可以使用 /sid 指令获取在某个平台上的会话 ID。也可在 AstrBot 日志内获取会话 ID，当一条消息没通过白名单时，会输出 INFO 级别的日志。会话 ID 类似 aiocqhttp:GroupMessage:547540978",
                    },
                    "id_whitelist_log": {
                        "description": "打印白名单日志",
@@ -228,6 +266,7 @@ CONFIG_METADATA_2 = {
                    "path_mapping": {
                        "description": "路径映射",
                        "type": "list",
+                        "items": {"type": "string"},
                        "obvious_hint": True,
                        "hint": "此功能解决由于文件系统不一致导致路径不存在的问题。格式为 <原路径>:<映射路径>。如 `/app/.config/QQ:/var/lib/docker/volumes/xxxx/_data`。这样，当消息平台下发的事件中图片和语音路径以 `/app/.config/QQ` 开头时，开头被替换为 `/var/lib/docker/volumes/xxxx/_data`。这在 AstrBot 或者平台协议端使用 Docker 部署时特别有用。",
                    }
@@ -286,7 +325,7 @@ CONFIG_METADATA_2 = {
                        "type": "openai_chat_completion",
                        "enable": True,
                        "key": [],
-                        "api_base": "",
+                        "api_base": "https://api.openai.com/v1",
                        "model_config": {
                            "model": "gpt-4o-mini",
                        },
@@ -341,6 +380,16 @@ CONFIG_METADATA_2 = {
                            "model": "glm-4-flash",
                        },
                    },
+                    "硅基流动": {
+                        "id": "siliconflow",
+                        "type": "openai_chat_completion",
+                        "enable": True,
+                        "key": [],
+                        "api_base": "https://api.siliconflow.cn/v1",
+                        "model_config": {
+                            "model": "deepseek-ai/DeepSeek-V3",
+                        },
+                    },
                    "llmtuner": {
                        "id": "llmtuner_default",
                        "type": "llm_tuner",
@@ -382,9 +431,17 @@ CONFIG_METADATA_2 = {
                        "api_key": "",
                        "api_base": "",
                        "model": "tts-1",
+                        "openai-tts-voice": "alloy",
+                        "timeout": "20",
                    },
                },
                "items": {
+                    "openai-tts-voice": {
+                        "description": "voice",
+                        "type": "string",
+                        "obvious_hint": True,
+                        "hint": "OpenAI TTS 的声音。OpenAI 默认支持：'alloy', 'echo', 'fable', 'onyx', 'nova', 'shimmer'",
+                    },
                    "whisper_hint": {
                        "description": "本地部署 Whisper 模型须知",
                        "type": "string",
@@ -415,7 +472,7 @@ CONFIG_METADATA_2 = {
                    "api_base": {
                        "description": "API Base URL",
                        "type": "string",
-                        "hint": "API Base URL 请在在模型提供商处获得。如使用时出现了 404 报错，可以尝试在地址末尾加上 `/v1`。",
+                        "hint": "API Base URL 请在在模型提供商处获得。如出现 404 报错，尝试在地址末尾加上 /v1",
                        "obvious_hint": True,
                    },
                    "base_model_path": {
@@ -501,16 +558,25 @@ CONFIG_METADATA_2 = {
                    "web_search": {
                        "description": "启用网页搜索",
                        "type": "bool",
-                        "hint": "能访问 Google 时效果最佳。如果 Google 访问失败，程序会依次访问 Bing, Sogo 搜索引擎。",
+                        "obvious_hint": True,
+                        "hint": "能访问 Google 时效果最佳（国内需要在 `其他配置` 开启 HTTP 代理）。如果 Google 访问失败，程序会依次访问 Bing, Sogo 搜索引擎。",
+                    },
+                    "web_search_link": {
+                        "description": "网页搜索引用链接",
+                        "type": "bool",
+                        "obvious_hint": True,
+                        "hint": "开启后，将会传入网页搜索结果的链接给模型，并引导模型输出引用链接。",
                    },
                    "identifier": {
                        "description": "启动识别群员",
                        "type": "bool",
+                        "obvious_hint": True,
                        "hint": "在 Prompt 前加上群成员的名字以让模型更好地了解群聊状态。启用将略微增加 token 开销。",
                    },
                    "datetime_system_prompt": {
                        "description": "启用日期时间系统提示",
                        "type": "bool",
+                        "obvious_hint": True,
                        "hint": "启用后，会在系统提示词中加上当前机器的日期时间。",
                    },
                    "default_personality": {
@@ -552,15 +618,15 @@ CONFIG_METADATA_2 = {
                    "begin_dialogs": {
                        "description": "预设对话",
                        "type": "list",
-                        "items": {},
-                        "hint": "可选。在每个对话前会插入这些预设对话。格式要求：第一句为用户，第二句为助手，以此类推。",
+                        "items": {"type": "string"},
+                        "hint": "可选。在每个对话前会插入这些预设对话。对话需要成对(用户和助手)，输入完一个角色的内容之后按【回车】。需要偶数个对话",
                        "obvious_hint": True,
                    },
                    "mood_imitation_dialogs": {
                        "description": "对话风格模仿",
                        "type": "list",
-                        "items": {},
-                        "hint": "旨在让模型尽可能模仿学习到所填写的对话的语气风格。格式和 `预设对话` 一样。",
+                        "items": {"type": "string"},
+                        "hint": "旨在让模型尽可能模仿学习到所填写的对话的语气风格。格式和 `预设对话` 一致。对话需要成对(用户和助手)，输入完一个角色的内容之后按【回车】。需要偶数个对话",
                        "obvious_hint": True,
                    },
                },
@@ -599,6 +665,70 @@ CONFIG_METADATA_2 = {
                    },
                },
            },
+            "provider_ltm_settings": {
+                "description": "聊天记忆增强(Beta)",
+                "type": "object",
+                "items": {
+                    "group_icl_enable": {
+                        "description": "群聊内记录各群员对话",
+                        "type": "bool",
+                        "obvious_hint": True,
+                        "hint": "启用后，会记录群聊内各群员的对话。使用 /reset 命令清除记录。推荐使用 gpt-4o-mini 模型。",
+                    },
+                    "group_message_max_cnt": {
+                        "description": "群聊消息最大数量",
+                        "type": "int",
+                        "obvious_hint": True,
+                        "hint": "群聊消息最大数量。超过此数量后，会自动清除旧消息。",
+                    },
+                    "image_caption": {
+                        "description": "启用图像转述(需要模型支持)",
+                        "type": "bool",
+                        "obvious_hint": True,
+                        "hint": "启用后，当接收到图片消息时，会使用模型先将图片转述为文字再进行后续处理。推荐使用 gpt-4o-mini 模型。",
+                    },
+                    "image_caption_prompt": {
+                        "description": "图像转述提示词",
+                        "type": "string"
+                    },
+                    "active_reply": {
+                        "description": "主动回复",
+                        "type": "object",
+                        "items": {
+                            "enable": {
+                                "description": "启用主动回复",
+                                "type": "bool",
+                                "obvious_hint": True,
+                                "hint": "启用后，会根据触发概率主动回复群聊内的对话。QQ官方API(qq_official)不可用",
+                            },
+                            "method": {
+                                "description": "回复方法",
+                                "type": "string",
+                                "options": ["possibility_reply"],
+                                "hint": "回复方法。possibility_reply 为根据概率回复",
+                            },
+                            "possibility_reply": {
+                                "description": "回复概率",
+                                "type": "float",
+                                "obvious_hint": True,
+                                "hint": "回复概率。当回复方法为 possibility_reply 时有效。当概率 >= 1 时，每条消息都会回复。",
+                            },
+                            "prompt": {
+                                "description": "提示词",
+                                "type": "string",
+                                "obvious_hint": True,
+                                "hint": "提示词。当提示词为空时，如果触发回复，prompt是触发的消息的内容；否则是提示词。此项可以和定时回复（暂未实现）配合使用。",
+                            },
+                        },
+                    },
+                    "put_history_to_prompt": {
+                        "description": "将群聊历史记录作为 prompt",
+                        "type": "bool",
+                        "obvious_hint": True,
+                        "hint": "需要先启用 group_icl_enable。此功能会将群聊历史记录放到 prompt 再请求。如果关闭，则是放在 system_prompt。如果开启了主动回复，建议启用，模型能够更好地完成回复任务。",
+                    }
+                },
+            },
        },
    },
    "misc_config_group": {
@@ -608,7 +738,8 @@ CONFIG_METADATA_2 = {
                "description": "机器人唤醒前缀",
                "type": "list",
                "items": {"type": "string"},
-                "hint": "在不 @ 机器人的情况下，可以通过外加消息前缀来唤醒机器人。",
+                "obvious_hint": True,
+                "hint": "在不 @ 机器人的情况下，可以通过外加消息前缀来唤醒机器人。更改此配置将影响整个 Bot 的功能唤醒，包括所有指令。如果您不保留 `/`，则内置指令（help等）将需要通过您的唤醒前缀来触发。",
            },
            "t2i": {
                "description": "文本转图像",
@@ -618,7 +749,7 @@ CONFIG_METADATA_2 = {
            "admins_id": {
                "description": "管理员 ID",
                "type": "list",
-                "items": {"type": "int"},
+                "items": {"type": "string"},
                "hint": "管理员 ID 列表，管理员可以使用一些特权命令，如 `update`, `plugin` 等。ID 可以通过 `/myid` 指令获得。回车添加，可添加多个。",
            },
            "http_proxy": {
@@ -25,9 +25,8 @@ class AstrBotCoreLifecycle:
        self.astrbot_config = astrbot_config
        self.db = db
        
-        if self.astrbot_config['http_proxy']:
-            os.environ['https_proxy'] = self.astrbot_config['http_proxy']
-            os.environ['http_proxy'] = self.astrbot_config['http_proxy']
+        os.environ['https_proxy'] = self.astrbot_config['http_proxy']
+        os.environ['http_proxy'] = self.astrbot_config['http_proxy']
    
    async def initialize(self):
        logger.info("AstrBot v"+ VERSION)
@@ -306,7 +306,7 @@ class Image(BaseMessageComponent):

 class Reply(BaseMessageComponent):
    type: ComponentType = "Reply"
-    id: int
+    id: T.Union[str, int]
    text: T.Optional[str] = ""
    qq: T.Optional[int] = 0
    time: T.Optional[int] = 0
@@ -13,12 +13,10 @@ class MessageChain():
    Attributes:
        `chain` (list): 用于顺序存储各个组件。
        `use_t2i_` (bool): 用于标记是否使用文本转图片服务。默认为 None，即跟随用户的设置。当设置为 True 时，将会使用文本转图片服务。
-        `is_split_` (bool): 用于标记是否分条发送消息。默认为 False。启用后，将会依次发送 chain 中的每个 component。
    '''
    
    chain: List[BaseMessageComponent] = field(default_factory=list)
    use_t2i_: Optional[bool] = None # None 为跟随用户设置
-    is_split_: Optional[bool] = False # 是否将消息分条发送。默认为 False。启用后，将会依次发送 chain 中的每个 component。
    
    def message(self, message: str):
        '''添加一条文本消息到消息链 `chain` 中。
@@ -77,16 +75,6 @@ class MessageChain():
        '''
        self.use_t2i_ = use_t2i
        return self
-    
-    def is_split(self, is_split: bool):
-        '''设置是否分条发送消息。默认为 False。启用后，将会依次发送 chain 中的每个 component。
-        
-        Note:
-            具体的效果以各适配器实现为准。
-            
-        '''
-        self.is_split_ = is_split
-        return self

 class EventResultType(enum.Enum):
    '''用于描述事件处理的结果类型。
@@ -113,7 +101,6 @@ class MessageEventResult(MessageChain):
    Attributes:
        `chain` (list): 用于顺序存储各个组件。
        `use_t2i_` (bool): 用于标记是否使用文本转图片服务。默认为 None，即跟随用户的设置。当设置为 True 时，将会使用文本转图片服务。
-        `is_split_` (bool): 用于标记是否分条发送消息。默认为 False。启用后，将会依次发送 chain 中的每个 component。
        `result_type` (EventResultType): 事件处理的结果类型。
    '''
    
@@ -153,5 +140,10 @@ class MessageEventResult(MessageChain):
        '''
        return self.result_content_type == ResultContentType.LLM_RESULT
    
+    def get_plain_text(self) -> str:
+        '''获取纯文本消息。这个方法将获取所有 Plain 组件的文本并拼接成一条消息。空格分隔。
+        '''
+        return " ".join([comp.text for comp in self.chain if isinstance(comp, Plain)])
+    
    
 CommandResult = MessageEventResult
@@ -21,6 +21,10 @@ class DifyRequestSubStage(Stage):
        req: ProviderRequest = None
        
        provider = self.ctx.plugin_manager.context.get_using_provider()
+        
+        if not provider:
+            return
+        
        if provider.meta().type != "dify":
            return
        
@@ -68,12 +68,23 @@ class LLMRequestSubStage(Stage):
            if _nested:
                req.func_tool = None # 暂时不支持递归工具调用
            llm_response = await provider.text_chat(**req.__dict__) # 请求 LLM
+            
+            # 执行 LLM 响应后的事件。
+            handlers = star_handlers_registry.get_handlers_by_event_type(EventType.OnLLMResponseEvent)
+            for handler in handlers:
+                try:
+                    await handler.handler(event, llm_response)
+                except BaseException:
+                    logger.error(traceback.format_exc())
+            
            await Metric.upload(llm_tick=1, model_name=provider.get_model(), provider_type=provider.meta().type)

            if llm_response.role == 'assistant':
                # text completion
                event.set_result(MessageEventResult().message(llm_response.completion_text)
                                 .set_result_content_type(ResultContentType.LLM_RESULT))
+            elif llm_response.role == 'err':
+                event.set_result(MessageEventResult().message(f"AstrBot 请求失败。\n错误信息: {llm_response.completion_text}"))
            elif llm_response.role == 'tool':
                # function calling
                function_calling_result = {}
@@ -39,8 +39,11 @@ class StarRequestSubStage(Stage):
            except Exception as e:
                logger.error(traceback.format_exc())
                logger.error(f"Star {handler.handler_full_name} handle error: {e}")
-                ret = f":(\n\n在调用插件 {star_map.get(handler.handler_module_path).name} 的处理函数 {handler.handler_name} 时出现异常：{e}"
-                event.set_result(MessageEventResult().message(ret))
-                yield
-                event.clear_result()
+                
+                if event.is_at_or_wake_command:
+                    ret = f":(\n\n在调用插件 {star_map.get(handler.handler_module_path).name} 的处理函数 {handler.handler_name} 时出现异常：{e}"
+                    event.set_result(MessageEventResult().message(ret))
+                    yield
+                    event.clear_result()
+                
                event.stop_event()
@@ -37,7 +37,11 @@ class ProcessStage(Stage):
                    # Handler 的 LLM 请求
                    logger.debug(f"llm request -> {resp.prompt}")
                    event.set_extra("provider_request", resp)
+                    _t = False
                    async for _ in self.llm_request_sub_stage.process(event):
+                        _t = True
+                        yield
+                    if not _t:
                        yield
                else:
                    yield
@@ -49,6 +53,11 @@ class ProcessStage(Stage):
        if not event._has_send_oper and event.is_at_or_wake_command:
            if (event.get_result() and not event.get_result().is_stopped()) or not event.get_result():
                provider = self.ctx.plugin_manager.context.get_using_provider()
+                
+                if not provider:
+                    logger.info("未找到可用的 LLM 提供商，请先前往配置服务提供商。")
+                    return
+                
                match provider.meta().type:
                    case "dify":
                        async for _ in self.dify_request_sub_stage.process(event):
@@ -1,7 +1,10 @@
+import random
+import asyncio
 from typing import Union, AsyncGenerator
 from ..stage import register_stage, Stage
 from ..context import PipelineContext
 from astrbot.core.platform.astr_message_event import AstrMessageEvent
+from astrbot.core.message.message_event_result import MessageChain
 from astrbot.core import logger
 from astrbot.core.star.star_handler import star_handlers_registry, EventType

@@ -9,6 +12,19 @@ from astrbot.core.star.star_handler import star_handlers_registry, EventType
 class RespondStage(Stage):
    async def initialize(self, ctx: PipelineContext):
        self.ctx = ctx
+        
+        # 分段回复
+        self.enable_seg: bool = ctx.astrbot_config['platform_settings']['segmented_reply']['enable']
+        self.only_llm_result = ctx.astrbot_config['platform_settings']['segmented_reply']['only_llm_result']
+        interval_str: str = ctx.astrbot_config['platform_settings']['segmented_reply']['interval']
+        interval_str_ls = interval_str.replace(" ", "").split(",")
+        try:
+            self.interval = [float(t) for t in interval_str_ls]
+        except BaseException as e:
+            logger.error(f'解析分段回复的间隔时间失败。{e}')
+            self.interval = [1.5, 3.5]
+        logger.info(f"分段回复间隔时间：{self.interval}")
+

    async def process(self, event: AstrMessageEvent) -> Union[None, AsyncGenerator[None, None]]:
        result = event.get_result()
@@ -16,7 +32,16 @@ class RespondStage(Stage):
            return

        if len(result.chain) > 0:
-            await event.send(result)
+            await event._pre_send()
+            
+            if self.enable_seg and ((self.only_llm_result and result.is_llm_result()) or not self.only_llm_result):
+                # 分段回复
+                for comp in result.chain:
+                    await event.send(MessageChain([comp]))
+                    await asyncio.sleep(random.uniform(self.interval[0], self.interval[1]))
+            else:
+                await event.send(result)
+            await event._post_send()
            logger.info(f"AstrBot -> {event.get_sender_name()}/{event.get_sender_id()}: {event._outline_chain(result.chain)}")
        
        handlers = star_handlers_registry.get_handlers_by_event_type(EventType.OnAfterMessageSentEvent)
@@ -1,4 +1,5 @@
 import time
+import re
 import traceback
 from typing import Union, AsyncGenerator
 from ..stage import register_stage
@@ -18,7 +19,11 @@ class ResultDecorateStage:
        self.reply_with_mention = ctx.astrbot_config['platform_settings']['reply_with_mention']
        self.reply_with_quote = ctx.astrbot_config['platform_settings']['reply_with_quote']
        self.use_tts = ctx.astrbot_config['provider_tts_settings']['enable']
-        self.t2i = ctx.astrbot_config['t2i']
+        
+        # 分段回复
+        self.enable_segmented_reply = ctx.astrbot_config['platform_settings']['segmented_reply']['enable']
+        self.only_llm_result = ctx.astrbot_config['platform_settings']['segmented_reply']['only_llm_result']
+        self.regex = ctx.astrbot_config['platform_settings']['segmented_reply']['regex']

    async def process(self, event: AstrMessageEvent) -> Union[None, AsyncGenerator[None, None]]:
        result = event.get_result()
@@ -33,29 +38,53 @@ class ResultDecorateStage:
        if len(result.chain) > 0:
            # 回复前缀
            if self.reply_prefix:
-                result.chain.insert(0, Plain(self.reply_prefix))
+                for comp in result.chain:
+                    if isinstance(comp, Plain):
+                        comp.text = self.reply_prefix + comp.text
+                        break
+            
+             # 分段回复 
+            if self.enable_segmented_reply:
+                if (self.only_llm_result and result.is_llm_result()) or not self.only_llm_result:
+                    new_chain = []
+                    for comp in result.chain:
+                        if isinstance(comp, Plain):
+                            split_response = re.findall(r".*?[。？！~…]+|.+$", comp.text)
+                            if not split_response:
+                                new_chain.append(comp)
+                                continue
+                            for seg in split_response:
+                                new_chain.append(Plain(seg))
+                        else:
+                            # 非 Plain 类型的消息段不分段
+                            new_chain.append(comp)
+                    result.chain = new_chain
                
            # TTS
            if self.use_tts and result.is_llm_result():
                tts_provider = self.ctx.plugin_manager.context.provider_manager.curr_tts_provider_inst
-                plain_str = ""
+                new_chain = []
                for comp in result.chain:
-                    if isinstance(comp, Plain):
-                        plain_str += " " + comp.text
+                    if isinstance(comp, Plain) and len(comp.text) > 1:
+                        try:
+                            logger.info("TTS 请求: " + comp.text)
+                            audio_path = await tts_provider.get_audio(comp.text)
+                            logger.info("TTS 结果: " + audio_path)
+                            if audio_path:
+                                new_chain.append(Record(file=audio_path, url=audio_path))
+                            else:
+                                logger.error(f"由于 TTS 音频文件没找到，消息段转语音失败: {comp.text}")
+                                new_chain.append(comp)
+                        except BaseException:
+                            traceback.print_exc()
+                            logger.error("TTS 失败，使用文本发送。")
+                            new_chain.append(comp)
                    else:
-                        break
-                if plain_str:
-                    try:
-                        audio_path = await tts_provider.get_audio(plain_str)
-                        logger.info("TTS 结果: " + audio_path)
-                        if audio_path:
-                            result.chain = [Record(file=audio_path, url=audio_path)]
-                    except BaseException:
-                        traceback.print_exc()
-                        logger.error("TTS 失败，使用文本发送。")
+                        new_chain.append(comp)
+                result.chain = new_chain
            
            # 文本转图片
-            elif (result.use_t2i_ is None and self.t2i) or result.use_t2i_:
+            elif (result.use_t2i_ is None and self.ctx.astrbot_config['t2i']) or result.use_t2i_:
                plain_str = ""
                for comp in result.chain:
                    if isinstance(comp, Plain):
@@ -74,8 +103,12 @@ class ResultDecorateStage:
                    if url:
                        result.chain = [Image.fromURL(url)]
            
+            # at 回复
            if self.reply_with_mention and event.get_message_type() != MessageType.FRIEND_MESSAGE:
-                result.chain.insert(0, At(qq=event.get_sender_id()))
+                result.chain.insert(0, At(qq=event.get_sender_id(), name=event.get_sender_name()))
+                if len(result.chain) > 1 and isinstance(result.chain[1], Plain):
+                    result.chain[1].text = "\n" + result.chain[1].text
            
+            # 引用回复
            if self.reply_with_quote:
                result.chain.insert(0, Reply(id=event.message_obj.message_id))
@@ -18,6 +18,11 @@ class WhitelistCheckStage(Stage):
        
    async def process(self, event: AstrMessageEvent) -> Union[None, AsyncGenerator[None, None]]:
        if not self.enable_whitelist_check:
+            # 白名单检查未启用
+            return
+        
+        if len(self.whitelist) == 0:
+            # 白名单为空，不检查
            return
        
        if event.get_platform_name() == 'webchat':
@@ -179,6 +179,15 @@ class AstrMessageEvent(abc.ABC):
        await Metric.upload(msg_event_tick = 1, adapter_name = self.platform_meta.name)
        self._has_send_oper = True
        
+    async def _pre_send(self):
+        '''调度器会在执行 send() 前调用该方法'''
+        pass
+    
+    async def _post_send(self):
+        '''调度器会在执行 send() 后调用该方法'''
+        pass
+        
+        
    def set_result(self, result: Union[MessageEventResult, str]):
        '''设置消息事件的结果。
        
@@ -287,6 +296,7 @@ class AstrMessageEvent(abc.ABC):
    def request_llm(
        self,
        prompt: str,
+        func_tool_manager = None,
        session_id: str = None,
        image_urls: List[str] = None,
        contexts: List = None,
@@ -302,11 +312,13 @@ class AstrMessageEvent(abc.ABC):
        
        image_urls: 可以是 base64:// 或者 http:// 开头的图片链接，也可以是本地图片路径。
        contexts: 当指定 contexts 时，将会**只**使用 contexts 作为上下文。
+        func_tool_manager: 函数工具管理器，用于调用函数工具。用 self.context.get_llm_tool_manager() 获取。
        '''
        return ProviderRequest(
            prompt = prompt,
            session_id = session_id,
            image_urls = image_urls,
+            func_tool = func_tool_manager,
            contexts = contexts,
            system_prompt = system_prompt
        )
@@ -24,7 +24,10 @@ class PlatformManager():
                case "qq_official":
                    from .sources.qqofficial.qqofficial_platform_adapter import QQOfficialPlatformAdapter # noqa: F401
                case "vchat":
-                    from .sources.vchat.vchat_platform_adapter import VChatPlatformAdapter # noqa: F401
+                    try:
+                        from .sources.vchat.vchat_platform_adapter import VChatPlatformAdapter # noqa: F401
+                    except BaseException:
+                        logger.warning("当前 astrbot 已不维护 vchat 的接入，如有需要请 pip 安装 vchat 然后重启")
                case "gewechat":
                    from .sources.gewechat.gewechat_platform_adapter import GewechatPlatformAdapter # noqa: F401
        
@@ -40,11 +40,5 @@ class AiocqhttpMessageEvent(AstrMessageEvent):
        ret = await AiocqhttpMessageEvent._parse_onebot_json(message)
        if os.environ.get('TEST_MODE', 'off') == 'on':
            return
-        
-        if message.is_split_: # 分条发送
-            for m in ret:
-                await self.bot.send(self.message_obj.raw_message, [m])
-                await asyncio.sleep(random.uniform(0.75, 2.5))
-        else:
-            await self.bot.send(self.message_obj.raw_message, ret)
+        await self.bot.send(self.message_obj.raw_message, ret)
        await super().send(message)
@@ -66,6 +66,7 @@ class SimpleGewechatClient():
        if type_name == "Offline":
            logger.critical("收到 gewechat 下线通知。")
            return
+        
        abm = AstrBotMessage()
        d = data['Data']
    
@@ -94,6 +95,8 @@ class SimpleGewechatClient():
            if f'<atuserlist><![CDATA[,{abm.self_id}]]>' in msg_source \
                or f'<atuserlist><![CDATA[{abm.self_id}]]>' in msg_source:
                at_me = True
+            if '在群聊中@了你' in d.get('PushContent', ''):
+                at_me = True
        else:
            abm.type = MessageType.FRIEND_MESSAGE
            user_id = from_user_name
@@ -102,7 +105,7 @@ class SimpleGewechatClient():
        if at_me:
            abm.message.insert(0, At(qq=abm.self_id))
        
-        user_real_name = d['PushContent'].split(' : ')[0] \
+        user_real_name = d.get('PushContent', 'unknown : ').split(' : ')[0] \
            .replace('在群聊中@了你', '') \
            .replace('在群聊中发了一段语音', '') # 真实昵称
        abm.sender = MessageMember(user_id, user_real_name)
@@ -153,13 +156,17 @@ class SimpleGewechatClient():
        if data.get('testMsg', None):
            return quart.jsonify({"r": "AstrBot ACK"})
        
-        abm = await self._convert(data)
-        
+        abm = None
+        try:
+            abm = await self._convert(data)
+        except BaseException as e:
+            logger.warning(f"尝试解析 GeweChat 下发的消息时遇到问题: {e}。下发消息内容: {data}。")
+            
        if abm:
            coro = getattr(self, "on_event_received")
            if coro:
                await coro(abm)
-            
+        
        return quart.jsonify({"r": "AstrBot ACK"})
    
    async def handle_file(self, file_id):
@@ -289,7 +296,7 @@ class SimpleGewechatClient():
            await asyncio.sleep(5)
            
        if appid:
-            sp.put(f"gewechat-appid-{nickname}", appid)
+            sp.put(f"gewechat-appid-{self.nickname}", appid)
            self.appid = appid
            logger.info(f"已保存 APPID: {appid}")
    
@@ -5,7 +5,7 @@ import botpy.types.message
 from astrbot.core.utils.io import file_to_base64, download_image_by_url
 from astrbot.api.event import AstrMessageEvent, MessageChain
 from astrbot.api.platform import AstrBotMessage, PlatformMetadata
-from astrbot.api.message_components import Plain, Image
+from astrbot.api.message_components import Plain, Image, Reply
 from botpy import Client
 from botpy.http import Route

@@ -14,12 +14,33 @@ class QQOfficialMessageEvent(AstrMessageEvent):
    def __init__(self, message_str: str, message_obj: AstrBotMessage, platform_meta: PlatformMetadata, session_id: str, bot: Client):
        super().__init__(message_str, message_obj, platform_meta, session_id)
        self.bot = bot
+        self.send_buffer = None
        
    async def send(self, message: MessageChain):
+        if not self.send_buffer:
+            self.send_buffer = message
+        else:
+            self.send_buffer.chain.extend(message.chain)
+
+    async def _post_send(self):
+        '''QQ 官方 API 仅支持回复一次'''
        source = self.message_obj.raw_message
        assert isinstance(source, (botpy.message.Message, botpy.message.GroupMessage, botpy.message.DirectMessage, botpy.message.C2CMessage))
        
-        plain_text, image_base64, image_path = await QQOfficialMessageEvent._parse_to_qqofficial(message)
+        plain_text, image_base64, image_path = await QQOfficialMessageEvent._parse_to_qqofficial(self.send_buffer)
+        
+        ref = None
+        for i in self.send_buffer.chain:
+            if isinstance(i, Reply):
+                try:
+                    ref = self.message_obj.raw_message.message_reference
+                    ref = botpy.types.message.Reference(
+                        message_id=ref.message_id,
+                        ignore_get_message_error=False
+                    )
+                except BaseException as _:
+                    pass
+                break
        
        payload = {
            'content': plain_text,
@@ -28,27 +49,37 @@ class QQOfficialMessageEvent(AstrMessageEvent):
        
        match type(source):
            case botpy.message.GroupMessage:
+                if ref:
+                    payload['message_reference'] = ref
                if image_base64:
                    media = await self.upload_group_and_c2c_image(image_base64, 1, group_openid=source.group_openid)
                    payload['media'] = media
                    payload['msg_type'] = 7
                await self.bot.api.post_group_message(group_openid=source.group_openid, **payload)
            case botpy.message.C2CMessage:
+                if ref:
+                    payload['message_reference'] = ref
                if image_base64:
                    media = await self.upload_group_and_c2c_image(image_base64, 1, openid=source.author.user_openid)
                    payload['media'] = media
                    payload['msg_type'] = 7
                await self.bot.api.post_c2c_message(openid=source.author.user_openid, **payload)
            case botpy.message.Message:
+                if ref:
+                    payload['message_reference'] = ref
                if image_path:
                    payload['file_image'] = image_path
                await self.bot.api.post_message(channel_id=source.channel_id, **payload)
            case botpy.message.DirectMessage:
+                if ref:
+                    payload['message_reference'] = ref
                if image_path:
                    payload['file_image'] = image_path
                await self.bot.api.post_dms(guild_id=source.guild_id, **payload)

-        await super().send(message)
+        await super().send(self.send_buffer)
+        
+        self.send_buffer = None
            
    async def upload_group_and_c2c_image(self, image_base64: str, file_type: int, **kwargs) -> botpy.types.message.Media:
        payload = {
@@ -2,6 +2,7 @@ import enum
 from dataclasses import dataclass, field
 from typing import List, Dict, Type
 from .func_tool_manager import FuncCall
+from openai.types.chat.chat_completion import ChatCompletion


 class ProviderType(enum.Enum):
@@ -51,4 +52,7 @@ class LLMResponse:
    tools_call_args: List[Dict[str, any]] = field(default_factory=list)
    '''工具调用参数'''
    tools_call_name: List[str] = field(default_factory=list)
-    '''工具调用名称'''
+    '''工具调用名称'''
+    
+    raw_completion: ChatCompletion = None
+    _new_record: Dict[str, any] = None
@@ -108,13 +108,19 @@ class FuncCall:
        for f in self.func_list:
            if not f.active:
                continue
-            tools.append(
-                {
-                    "name": f.name,
-                    "parameters": f.parameters,
-                    "description": f.description,
-                }
-            )
+
+            func_declaration = {
+                "name": f.name,
+                "description": f.description
+            }
+
+            # 检查并添加非空的properties参数
+            params = f.parameters if isinstance(f.parameters, dict) else {}
+            if params.get("properties", {}):
+                func_declaration["parameters"] = params
+
+            tools.append(func_declaration)
+
        declarations["function_declarations"] = tools
        return declarations
        
@@ -1,4 +1,5 @@
 import traceback
+import uuid
 from astrbot.core.config.astrbot_config import AstrBotConfig
 from .provider import Provider, STTProvider, TTSProvider, Personality
 from .entites import ProviderType
@@ -13,8 +14,11 @@ class ProviderManager():
        self.providers_config: List = config['provider']
        self.provider_settings: dict = config['provider_settings']
        self.provider_stt_settings: dict = config.get('provider_stt_settings', {})
+        self.provider_tts_settings: dict = config.get('provider_tts_settings', {})
        self.persona_configs: list = config.get('persona', [])
        
+        # 人格情景管理
+        # 目前没有拆成独立的模块
        self.default_persona_name = self.provider_settings.get('default_personality', 'default')
        self.personas: List[Personality] = []
        self.selected_default_persona = None
@@ -26,7 +30,7 @@ class ProviderManager():
            if begin_dialogs:
                if len(begin_dialogs) % 2 != 0:
                    logger.error(f"{persona['name']} 人格情景预设对话格式不对，条数应该为偶数。")
-                    continue
+                    begin_dialogs = []
                user_turn = True
                for dialog in begin_dialogs:
                    bd_processed.append({
@@ -38,9 +42,9 @@ class ProviderManager():
            if mood_imitation_dialogs:
                if len(mood_imitation_dialogs) % 2 != 0:
                    logger.error(f"{persona['name']} 对话风格对话格式不对，条数应该为偶数。")
-                    continue
+                    mood_imitation_dialogs = []
                user_turn = True
-                for dialog in begin_dialogs:
+                for dialog in mood_imitation_dialogs:
                    role = "A" if user_turn else "B"
                    mid_processed += f"{role}: {dialog}\n"
                    if not user_turn:
@@ -59,6 +63,10 @@ class ProviderManager():
            except Exception as e:
                logger.error(f"解析 Persona 配置失败：{e}")
        
+        if not self.selected_default_persona and len(self.personas) > 0:
+            # 默认选择第一个
+            self.selected_default_persona = self.personas[0]
+        
        
        self.provider_insts: List[Provider] = []
        '''加载的 Provider 的实例'''
@@ -82,12 +90,16 @@ class ProviderManager():
        if kdb_cfg and len(kdb_cfg):
            self.curr_kdb_name = list(kdb_cfg.keys())[0]
        
+        changed = False
        for provider_cfg in self.providers_config:
            if not provider_cfg['enable']:
                continue
            
            if provider_cfg['id'] in self.loaded_ids:
-                raise ValueError(f"Provider ID 重复：{provider_cfg['id']}。")
+                new_id = f"{provider_cfg['id']}_{str(uuid.uuid4())[:8]}"
+                logger.info(f"Provider ID 重复：{provider_cfg['id']}。已自动更改为 {new_id}。")
+                provider_cfg['id'] = new_id
+                changed = True
            self.loaded_ids[provider_cfg['id']] = True
            
            try:
@@ -115,21 +127,29 @@ class ProviderManager():
            except Exception as e:
                logger.critical(f"加载 {provider_cfg['type']}({provider_cfg['id']}) 提供商适配器失败：{e}。未知原因")
                continue
-            
+        
+        if changed:
+            try:
+                config.save_config()
+            except Exception as e:
+                logger.warning(f"保存配置文件失败：{e}")
+
    async def initialize(self):
+        
+        selected_provider_id = sp.get("curr_provider")
+        selected_stt_provider_id = self.provider_stt_settings.get("provider_id")
+        selected_tts_provider_id = self.provider_settings.get("provider_id")
+        provider_enabled = self.provider_settings.get("enable", False)
+        stt_enabled = self.provider_stt_settings.get("enable", False)
+        tts_enabled = self.provider_tts_settings.get("enable", False)
+            
        for provider_config in self.providers_config:
            if not provider_config['enable']:
                continue
            if provider_config['type'] not in provider_cls_map:
                logger.error(f"未找到适用于 {provider_config['type']}({provider_config['id']}) 的提供商适配器，请检查是否已经安装或者名称填写错误。已跳过。")
                continue
-            selected_provider_id = sp.get("curr_provider")
-            selected_stt_provider_id = self.provider_stt_settings.get("provider_id")
-            selected_tts_provider_id = self.provider_settings.get("provider_id")
-            provider_enabled = self.provider_settings.get("enable", False)
-            stt_enabled = self.provider_stt_settings.get("enable", False)
-            tts_enabled = self.provider_settings.get("enable", False)
-            
+
            provider_metadata = provider_cls_map[provider_config['type']]
            logger.info(f"尝试实例化 {provider_config['type']}({provider_config['id']}) 提供商适配器 ...")
            try:
@@ -74,7 +74,7 @@ class Provider(AbstractProvider):
        if persistant_history:
            # 读取历史记录
            try:
-                for history in db_helper.get_llm_history(provider_type=provider_config['type']):
+                for history in db_helper.get_llm_history(provider_type=provider_config['id']):
                    self.session_memory[history.session_id] = json.loads(history.content)
            except BaseException as e:
                logger.warning(f"读取 LLM 对话历史记录 失败：{e}。仍可正常使用。")
@@ -190,7 +190,8 @@ class ProviderGoogleGenAI(Provider):
                llm_response.role = "tool"
                llm_response.tools_call_args.append(candidate['functionCall']['args'])
                llm_response.tools_call_name.append(candidate['functionCall']['name'])
-            
+        
+        llm_response.completion_text = llm_response.completion_text.strip()
        return llm_response


@@ -260,10 +261,11 @@ class ProviderGoogleGenAI(Provider):
                    "role": "assistant",
                    "content": llm_response.completion_text
                }]
-            self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['type'])
+            self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['id'])
        
    async def forget(self, session_id: str) -> bool:
        self.session_memory[session_id] = []
+        self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['id'])
        return True

    def get_current_key(self) -> str:
@@ -118,10 +118,11 @@ class LLMTunerModelLoader(Provider):
                    "role": "assistant",
                    "content": llm_response.completion_text
                }]
-            self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['type'])
+            self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['id'])
        
    async def forget(self, session_id):
        self.session_memory[session_id] = []
+        self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['id'])
        return True

    async def get_current_key(self):
@@ -1,5 +1,6 @@
 import base64
 import json
+import os

 from openai import AsyncOpenAI, NOT_GIVEN
 from openai.types.chat.chat_completion import ChatCompletion
@@ -71,31 +72,21 @@ class ProviderOpenAIOfficial(Provider):
        except NotFoundError as e:
            raise Exception(f"获取模型列表失败：{e}")
    
-    async def pop_record(self, session_id: str, pop_system_prompt: bool = False):
+    async def pop_record(self, session_id: str):
        '''
-        弹出第一条记录
+        弹出最早的一个对话
        '''
        if session_id not in self.session_memory:
            raise Exception("会话 ID 不存在")

-        if len(self.session_memory[session_id]) == 0:
-            return None
-
-        for i in range(len(self.session_memory[session_id])):
-            # 检查是否是 system prompt
-            if not pop_system_prompt and self.session_memory[session_id][i]['user']['role'] == "system":
-                # 如果只有一个 system prompt，才不删掉
-                f = False
-                for j in range(i+1, len(self.session_memory[session_id])):
-                    if self.session_memory[session_id][j]['user']['role'] == "system":
-                        f = True
-                        break
-                if not f:
-                    continue
-            record = self.session_memory[session_id].pop(i)
-            break
-
-        return record
+        if len(self.session_memory[session_id]) < 2:
+            return
+        
+        try:
+            self.session_memory[session_id].pop(0)
+            self.session_memory[session_id].pop(0)
+        except IndexError:
+            pass
    
    async def _query(self, payloads: dict, tools: FuncCall) -> LLMResponse:
        if tools:
@@ -107,7 +98,7 @@ class ProviderOpenAIOfficial(Provider):
            **payloads,
            stream=False
        )
-        
+
        assert isinstance(completion, ChatCompletion)
        logger.debug(f"completion: {completion}")

@@ -118,7 +109,8 @@ class ProviderOpenAIOfficial(Provider):
        if choice.message.content:
            # text completion
            completion_text = str(choice.message.content).strip()
-            return LLMResponse("assistant", completion_text)
+
+            return LLMResponse("assistant", completion_text, raw_completion=completion)
        elif choice.message.tool_calls:
            # tools call (function calling)
            args_ls = []
@@ -129,8 +121,9 @@ class ProviderOpenAIOfficial(Provider):
                        args = json.loads(tool_call.function.arguments)
                        args_ls.append(args)
                        func_name_ls.append(tool_call.function.name)
-            return LLMResponse(role="tool", tools_call_args=args_ls, tools_call_name=func_name_ls)
+            return LLMResponse(role="tool", tools_call_args=args_ls, tools_call_name=func_name_ls, raw_completion=completion)
        else:
+            logger.error(f"API 返回的 completion 无法解析：{completion}。")
            raise Exception("Internal Error")

    async def text_chat(
@@ -160,18 +153,17 @@ class ProviderOpenAIOfficial(Provider):
            "messages": context_query,
            **self.provider_config.get("model_config", {})
        }
-
+        llm_response = None
        try:
            llm_response = await self._query(payloads, func_tool)
-            await self.save_history(contexts, new_record, session_id, llm_response)
-            return llm_response
        except Exception as e:
            if "maximum context length" in str(e):
+                # 重试 10 次
                retry_cnt = 10
                while retry_cnt > 0:
-                    logger.warning(f"请求失败：{e}。上下文长度超过限制。尝试弹出最早的记录然后重试。")
+                    logger.warning("上下文长度超过限制。尝试弹出最早的记录然后重试。")
                    try:
-                        self.pop_record(session_id)
+                        await self.pop_record(session_id)
                        llm_response = await self._query(payloads, func_tool)
                        break
                    except Exception as e:
@@ -179,8 +171,67 @@ class ProviderOpenAIOfficial(Provider):
                            retry_cnt -= 1
                        else:
                            raise e
+                if retry_cnt == 0:
+                    llm_response = LLMResponse("err", "err: 请尝试 /reset 清除会话记录。")
+            elif "The model is not a VLM" in str(e): # siliconcloud
+                # 尝试删除所有 image
+                print(context_query)
+                new_contexts = await self._remove_image_from_context(context_query)
+                print(new_contexts)
+                payloads['messages'] = new_contexts
+                llm_response = await self._query(payloads, func_tool)
+
+            elif 'does not support Function Calling' in str(e) \
+                or 'does not support tools' in str(e)  \
+                or 'Function call is not supported' in str(e) \
+                or 'Tool calling is not supported' in str(e): # siliconcloud 
+                    logger.info(f"{self.get_model()} 不支持函数调用工具调用，已经自动去除")
+                    if 'tools' in payloads:
+                        del payloads['tools']
+                    llm_response = await self._query(payloads, None)
            else:
+                logger.error(f"发生了错误。Provider 配置如下: {self.provider_config}")
+                
+                if 'tool' in str(e).lower() and 'support' in str(e).lower():
+                    logger.error(f"疑似该模型不支持函数调用工具调用。请输入 /tool off_all")
+                
+                if 'Connection error.' in str(e):
+                    proxy = os.environ.get("http_proxy", None)
+                    if proxy:
+                        logger.error(f"可能为代理原因，请检查代理是否正常。当前代理: {proxy}")
+                
                raise e
+        
+        if kwargs.get("persist", True) and llm_response:
+            await self.save_history(contexts, new_record, session_id, llm_response)
+        
+        return llm_response
+    
+    async def _remove_image_from_context(self, contexts: List):
+        '''
+        从上下文中删除所有带有 image 的记录
+        '''
+        new_contexts = []
+        
+        flag = False
+        for context in contexts:
+            if flag:
+                flag = False # 删除 image 后，下一条（LLM 响应）也要删除
+                continue
+            if isinstance(context['content'], list):
+                flag = True
+                # continue
+                new_content = []
+                for item in context['content']:
+                    if isinstance(item, dict) and 'image_url' in item:
+                        continue
+                    new_content.append(item)
+                if not new_content:
+                    # 用户只发了图片
+                    new_content = [{"type": "text", "text": "[图片]"}]
+                context['content'] = new_content
+            new_contexts.append(context)
+        return new_contexts

    
    async def save_history(self, contexts: List, new_record: dict, session_id: str, llm_response: LLMResponse):
@@ -200,10 +251,11 @@ class ProviderOpenAIOfficial(Provider):
                    "role": "assistant",
                    "content": llm_response.completion_text
                }]
-            self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['type'])
+            self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['id'])
        
    async def forget(self, session_id: str) -> bool:
        self.session_memory[session_id] = []
+        self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['id'])
        return True

    def get_current_key(self) -> str:
@@ -15,7 +15,7 @@ class ProviderOpenAITTSAPI(TTSProvider):
    ) -> None:
        super().__init__(provider_config, provider_settings)
        self.chosen_api_key = provider_config.get("api_key", "")
-        self.voice = provider_config.get("voice", "alloy")
+        self.voice = provider_config.get("openai-tts-voice", "alloy")

        self.client = AsyncOpenAI(
            api_key=self.chosen_api_key,
@@ -1,3 +1,7 @@
+'''
+此功能已过时，参考 https://astrbot.app/dev/plugin.html#%E6%B3%A8%E5%86%8C%E6%8F%92%E4%BB%B6%E9%85%8D%E7%BD%AE-beta
+'''
+
 from typing import Union
 import os
 import json
@@ -10,7 +10,7 @@ from astrbot.core.platform.astr_message_event import MessageSesion
 from astrbot.core.message.message_event_result import MessageChain
 from astrbot.core.provider.manager import ProviderManager
 from astrbot.core.platform.manager import PlatformManager
-from .star import star_registry, StarMetadata
+from .star import star_registry, StarMetadata, star_map
 from .star_handler import star_handlers_registry, StarHandlerMetadata, EventType
 from .filter.command import CommandFilter
 from .filter.regex import RegexFilter
@@ -54,46 +54,19 @@ class Context:
        self.knowledge_db_manager = knowledge_db_manager

    def get_registered_star(self, star_name: str) -> StarMetadata:
+        '''根据插件名获取插件的 Metadata'''
        for star in star_registry:
            if star.name == star_name:
                return star

    def get_all_stars(self) -> List[StarMetadata]:
+        '''获取当前载入的所有插件 Metadata 的列表'''
        return star_registry
    
    def get_llm_tool_manager(self) -> FuncCall:
-        '''
-        获取 LLM Tool Manager
-        '''
+        '''获取 LLM Tool Manager，其用于管理注册的所有的 Function-calling tools'''
        return self.provider_manager.llm_tools
        
-    def register_llm_tool(self, name: str, func_args: list, desc: str, func_obj: Awaitable) -> None:
-        '''
-        为函数调用（function-calling / tools-use）添加工具。
-        
-        @param name: 函数名
-        @param func_args: 函数参数列表，格式为 [{"type": "string", "name": "arg_name", "description": "arg_description"}, ...]
-        @param desc: 函数描述
-        @param func_obj: 异步处理函数。
-        
-        异步处理函数会接收到额外的的关键词参数：event: AstrMessageEvent, context: Context。
-        '''
-        md = StarHandlerMetadata(
-            event_type=EventType.OnLLMRequestEvent,
-            handler_full_name=func_obj.__module__ + "_" + func_obj.__name__,
-            handler_name=func_obj.__name__,
-            handler_module_path=func_obj.__module__,
-            handler=func_obj,
-            event_filters=[],
-            desc=desc
-        )
-        star_handlers_registry.append(md)
-        self.provider_manager.llm_tools.add_func(name, func_args, desc, func_obj, func_obj)
-    
-    def unregister_llm_tool(self, name: str) -> None:
-        '''删除一个函数调用工具。如果再要启用，需要重新注册。'''
-        self.provider_manager.llm_tools.remove_func(name)
-        
    def activate_llm_tool(self, name: str) -> bool:
        '''激活一个已经注册的函数调用工具。注册的工具默认是激活状态。
        
@@ -102,6 +75,11 @@ class Context:
        '''
        func_tool = self.provider_manager.llm_tools.get_func(name)
        if func_tool is not None:
+            
+            if func_tool.handler_module_path in star_map:
+                if not star_map[func_tool.handler_module_path].activated:
+                    raise ValueError(f"此函数调用工具所属的插件 {star_map[func_tool.handler_module_path].name} 已被禁用，请先在管理面板启用再激活此工具。")
+            
            func_tool.active = True
            
            inactivated_llm_tools: list = sp.get("inactivated_llm_tools", [])
@@ -129,6 +107,101 @@ class Context:
            return True
        return False
    
+    def register_provider(self, provider: Provider):
+        '''
+        注册一个 LLM Provider(Chat_Completion 类型)。
+        '''
+        self.provider_manager.provider_insts.append(provider)
+        
+    def get_provider_by_id(self, provider_id: str) -> Provider:
+        '''通过 ID 获取用于文本生成任务的 LLM Provider(Chat_Completion 类型)。'''
+        for provider in self.provider_manager.provider_insts:
+            if provider.meta().id == provider_id:
+                return provider
+        return None
+    
+    def get_all_providers(self) -> List[Provider]:
+        '''获取所有用于文本生成任务的 LLM Provider(Chat_Completion 类型)。'''
+        return self.provider_manager.provider_insts
+    
+    def get_using_provider(self) -> Provider:
+        '''
+        获取当前使用的用于文本生成任务的 LLM Provider(Chat_Completion 类型)。
+        
+        通过 /provider 指令切换。
+        '''
+        return self.provider_manager.curr_provider_inst
+    
+    def get_config(self) -> AstrBotConfig:
+        '''获取 AstrBot 的配置。'''
+        return self._config
+    
+    def get_db(self) -> BaseDatabase:
+        '''获取 AstrBot 数据库。'''
+        return self._db
+    
+    def get_event_queue(self) -> Queue:
+        '''
+        获取事件队列。
+        '''
+        return self._event_queue
+    
+    async def send_message(self, session: Union[str, MessageSesion], message_chain: MessageChain) -> bool:
+        '''
+        根据 session(unified_msg_origin) 发送消息。
+        
+        @param session: 消息会话。通过 event.session 或者 event.unified_msg_origin 获取。
+        @param message_chain: 消息链。
+        
+        @return: 是否找到匹配的平台。
+        
+        当 session 为字符串时，会尝试解析为 MessageSesion 对象，如果解析失败，会抛出 ValueError 异常。
+        '''
+        
+        if isinstance(session, str):
+            try:
+                session = MessageSesion.from_str(session)
+            except BaseException as e:
+                raise ValueError("不合法的 session 字符串: " + str(e))
+
+        for platform in self.platform_manager.platform_insts:
+            if platform.meta().name == session.platform_name:
+                await platform.send_by_session(session, message_chain)
+                return True
+        return False
+
+    '''
+    以下的方法已经不推荐使用。请从 AstrBot 文档查看更好的注册方式。
+    '''
+    
+    def register_llm_tool(self, name: str, func_args: list, desc: str, func_obj: Awaitable) -> None:
+        '''
+        为函数调用（function-calling / tools-use）添加工具。
+        
+        @param name: 函数名
+        @param func_args: 函数参数列表，格式为 [{"type": "string", "name": "arg_name", "description": "arg_description"}, ...]
+        @param desc: 函数描述
+        @param func_obj: 异步处理函数。
+        
+        异步处理函数会接收到额外的的关键词参数：event: AstrMessageEvent, context: Context。
+        '''
+        md = StarHandlerMetadata(
+            event_type=EventType.OnLLMRequestEvent,
+            handler_full_name=func_obj.__module__ + "_" + func_obj.__name__,
+            handler_name=func_obj.__name__,
+            handler_module_path=func_obj.__module__,
+            handler=func_obj,
+            event_filters=[],
+            desc=desc
+        )
+        star_handlers_registry.append(md)
+        self.provider_manager.llm_tools.add_func(name, func_args, desc, func_obj, func_obj)
+    
+    def unregister_llm_tool(self, name: str) -> None:
+        '''删除一个函数调用工具。如果再要启用，需要重新注册。'''
+        self.provider_manager.llm_tools.remove_func(name)
+        
+
    def register_commands(self, star_name: str, command_name: str, desc: str, priority: int, awaitable: Awaitable, use_regex=False, ignore_prefix=False):
        '''
        注册一个命令。
@@ -162,77 +235,6 @@ class Context:
            ))
        star_handlers_registry.append(md)
    
-    def register_provider(self, provider: Provider):
-        '''
-        注册一个 LLM Provider(Chat_Completion 类型)。
-        '''
-        self.provider_manager.provider_insts.append(provider)
-        
-    def get_provider_by_id(self, provider_id: str) -> Provider:
-        '''
-        通过 ID 获取 LLM Provider(Chat_Completion 类型)。
-        '''
-        for provider in self.provider_manager.provider_insts:
-            if provider.meta().id == provider_id:
-                return provider
-        return None
-    
-    def get_all_providers(self) -> List[Provider]:
-        '''
-        获取所有 LLM Provider(Chat_Completion 类型)。
-        '''
-        return self.provider_manager.provider_insts
-    
-    def get_using_provider(self) -> Provider:
-        '''
-        获取当前使用的 LLM Provider(Chat_Completion 类型)。
-        
-        通过 /provider 指令切换。
-        '''
-        return self.provider_manager.curr_provider_inst
-    
-    def get_config(self) -> AstrBotConfig:
-        '''
-        获取 AstrBot 配置信息。
-        '''
-        return self._config
-    
-    def get_db(self) -> BaseDatabase:
-        '''
-        获取 AstrBot 数据库。
-        '''
-        return self._db
-    
-    def get_event_queue(self) -> Queue:
-        '''
-        获取事件队列。
-        '''
-        return self._event_queue
-    
-    async def send_message(self, session: Union[str, MessageSesion], message_chain: MessageChain) -> bool:
-        '''
-        根据 session(unified_msg_origin) 发送消息。
-        
-        @param session: 消息会话。通过 event.session 或者 event.unified_msg_origin 获取。
-        @param message_chain: 消息链。
-        
-        @return: 是否找到匹配的平台。
-        
-        当 session 为字符串时，会尝试解析为 MessageSesion 对象，如果解析失败，会抛出 ValueError 异常。
-        '''
-        
-        if isinstance(session, str):
-            try:
-                session = MessageSesion.from_str(session)
-            except BaseException as e:
-                raise ValueError("不合法的 session 字符串: " + str(e))
-
-        for platform in self.platform_manager.platform_insts:
-            if platform.meta().name == session.platform_name:
-                await platform.send_by_session(session, message_chain)
-                return True
-        return False
-
    def register_task(self, task: Awaitable, desc: str):
        '''
        注册一个异步任务。
@@ -7,6 +7,7 @@ from .star_handler import (
    register_regex,
    register_permission_type,
    register_on_llm_request,
+    register_on_llm_response,
    register_llm_tool,
    register_on_decorating_result,
    register_after_message_sent
@@ -21,6 +22,7 @@ __all__ = [
    'register_regex',
    'register_permission_type',
    'register_on_llm_request',
+    'register_on_llm_response',
    'register_llm_tool',
    'register_on_decorating_result',
    'register_after_message_sent'
@@ -139,6 +139,8 @@ def register_on_llm_request():
    
    Examples:
    ```py
+    from astrbot.api.provider import ProviderRequest
+    
    @on_llm_request()
    async def test(self, event: AstrMessageEvent, request: ProviderRequest) -> None:
        request.system_prompt += "你是一个猫娘..."
@@ -152,6 +154,27 @@ def register_on_llm_request():
    
    return decorator

+def register_on_llm_response():
+    '''当有 LLM 请求后的事件
+    
+    Examples:
+    ```py
+    from astrbot.api.provider import LLMResponse
+    
+    @on_llm_response()
+    async def test(self, event: AstrMessageEvent, response: LLMResponse) -> None:
+        ...
+    ```
+    
+    请务必接收两个参数：event, request
+    '''
+    def decorator(awaitable):
+        _ = get_handler_or_create(awaitable, EventType.OnLLMResponseEvent)
+        return awaitable
+    
+    return decorator
+
+
 def register_llm_tool(name: str = None):
    '''为函数调用（function-calling / tools-use）添加工具。
    
@@ -3,6 +3,7 @@ from __future__ import annotations
 from types import ModuleType
 from typing import List, Dict
 from dataclasses import dataclass
+from astrbot.core.config import AstrBotConfig

 star_registry: List[StarMetadata] = []
 star_map: Dict[str, StarMetadata] = {}
@@ -11,7 +12,7 @@ star_map: Dict[str, StarMetadata] = {}
@dataclass
 class StarMetadata:
    '''
-    Star 的元数据。
+    插件的元数据。
    '''
    name: str
    author: str  # 插件作者
@@ -20,21 +21,24 @@ class StarMetadata:
    repo: str = None  # 插件仓库地址

    star_cls_type: type = None
-    '''Star 的类对象的类型'''
+    '''插件的类对象的类型'''
    module_path: str = None
-    '''Star 的模块路径'''
+    '''插件的模块路径'''
    
    star_cls: object = None
-    '''Star 的类对象'''
+    '''插件的类对象'''
    module: ModuleType = None
-    '''Star 的模块对象'''
+    '''插件的模块对象'''
    root_dir_name: str = None
-    '''Star 的根目录名'''
+    '''插件的目录名称'''
    reserved: bool = False
-    '''是否是 AstrBot 的保留 Star'''
+    '''是否是 AstrBot 的保留插件'''
    
    activated: bool = True
    '''是否被激活'''
+    
+    config: AstrBotConfig = None
+    '''插件配置'''

    def __str__(self) -> str:
        return f"StarMetadata({self.name}, {self.desc}, {self.version}, {self.repo})"
@@ -47,6 +47,7 @@ class EventType(enum.Enum):
    '''
    AdapterMessageEvent = enum.auto() # 收到适配器发来的消息
    OnLLMRequestEvent = enum.auto() # 收到 LLM 请求（可以是用户也可以是插件）
+    OnLLMResponseEvent = enum.auto() # LLM 响应后
    OnDecoratingResultEvent = enum.auto() # 发送消息前
    OnCallingFuncToolEvent = enum.auto() # 调用函数工具
    OnAfterMessageSentEvent = enum.auto() # 发送消息后
@@ -2,12 +2,14 @@ import inspect
 import functools
 import os
 import sys
+import json
 import traceback
 import yaml
 import logging
 from types import ModuleType
 from typing import List
 from astrbot.core.config.astrbot_config import AstrBotConfig
+from astrbot.core.config.default import DEFAULT_VALUE_MAP
 from astrbot.core import logger, sp, pip_installer
 from .context import Context
 from . import StarMetadata
@@ -26,13 +28,20 @@ class PluginManager:
        self.updator = PluginUpdator(config['plugin_repo_mirror'])
        
        self.context = context
-        self.context._star_manager = self # 就这样吧，不想改了
+        self.context._star_manager = self
                
        self.config = config
        self.plugin_store_path = os.path.abspath(os.path.join(os.path.dirname(os.path.abspath(__file__)), "../../../data/plugins"))
+        '''存储插件的路径。即 data/plugins'''
+        self.plugin_config_path = os.path.abspath(os.path.join(os.path.dirname(os.path.abspath(__file__)), "../../../data/config"))
+        '''存储插件配置的路径。data/config'''
        self.reserved_plugin_path = os.path.abspath(os.path.join(os.path.dirname(os.path.abspath(__file__)), "../../../packages"))
+        '''保留插件的路径。在 packages 目录下'''
+        self.conf_schema_fname = "_conf_schema.json"
+        '''插件配置 Schema 文件名'''

    def _get_classes(self, arg: ModuleType):
+        '''获取指定模块（可以理解为一个 python 文件）下所有的类'''
        classes = []
        clsmembers = inspect.getmembers(arg, inspect.isclass)
        for (name, _) in clsmembers:
@@ -128,7 +137,7 @@ class PluginManager:
        return metadata
    
    async def reload(self):
-        '''扫描并加载所有的 Star'''
+        '''扫描并加载所有的插件'''
        for smd in star_registry:
            logger.debug(f"尝试终止插件 {smd.name} ...")
            if hasattr(smd.star_cls, "__del__"):
@@ -150,13 +159,13 @@ class PluginManager:
        inactivated_plugins: list = sp.get("inactivated_plugins", [])
        inactivated_llm_tools: list = sp.get("inactivated_llm_tools", [])
        
-        # 导入 Star 模块，并尝试实例化 Star 类
+        # 导入插件模块，并尝试实例化插件类
        for plugin_module in plugin_modules:
            try:
                module_str = plugin_module['module']
                # module_path = plugin_module['module_path']
-                root_dir_name = plugin_module['pname']
-                reserved = plugin_module.get('reserved', False)
+                root_dir_name = plugin_module['pname'] # 插件的目录名
+                reserved = plugin_module.get('reserved', False) # 是否是保留插件。目前在 packages/ 目录下的都是保留插件。保留插件不可以卸载。
                
                logger.info(f"正在载入插件 {root_dir_name} ...")
                
@@ -173,11 +182,33 @@ class PluginManager:
                    logger.error(traceback.format_exc())
                    logger.error(f"插件 {root_dir_name} 导入失败。原因：{str(e)}")
                    continue
+                
+                # 检查 _conf_schema.json
+                plugin_config = None
+                plugin_dir_path = os.path.join(self.plugin_store_path, root_dir_name) \
+                    if not reserved else os.path.join(self.reserved_plugin_path, root_dir_name)
+                plugin_schema_path = os.path.join(plugin_dir_path, self.conf_schema_fname)
+                if os.path.exists(plugin_schema_path):
+                    # 加载插件配置
+                    with open(plugin_schema_path, 'r', encoding='utf-8') as f:
+                        plugin_config = AstrBotConfig(
+                            config_path=os.path.join(self.plugin_config_path, f"{root_dir_name}_config.json"), 
+                            schema=json.loads(f.read())
+                        )

                if path in star_map:
                    # 通过装饰器的方式注册插件
                    metadata = star_map[path]
-                    metadata.star_cls = metadata.star_cls_type(context=self.context)
+                    
+                    if plugin_config:
+                        metadata.config = plugin_config
+                        try:
+                            metadata.star_cls = metadata.star_cls_type(context=self.context, config=plugin_config)
+                        except TypeError as _:
+                            metadata.star_cls = metadata.star_cls_type(context=self.context)
+                    else:
+                        metadata.star_cls = metadata.star_cls_type(context=self.context)
+                            
                    metadata.module = module
                    metadata.root_dir_name = root_dir_name
                    metadata.reserved = reserved
@@ -199,16 +230,20 @@ class PluginManager:
                    # v3.4.0 以前的方式注册插件
                    logger.debug(f"插件 {path} 未通过装饰器注册。尝试通过旧版本方式载入。")
                    classes = self._get_classes(module)
-                    try:
-                        obj = getattr(module, classes[0])(context=self.context)
-                    except BaseException as e:
-                        logger.error(f"插件 {root_dir_name} 实例化失败。")
-                        raise e
                    
+                    if plugin_config:
+                        try:
+                            obj = getattr(module, classes[0])(context=self.context, config=plugin_config) # 实例化插件类
+                        except TypeError as _:
+                            obj = getattr(module, classes[0])(context=self.context) # 实例化插件类
+                    else:
+                        obj = getattr(module, classes[0])(context=self.context) # 实例化插件类
+
                    metadata = None
                    plugin_path = os.path.join(self.plugin_store_path, root_dir_name) if not reserved else os.path.join(self.reserved_plugin_path, root_dir_name)
                    metadata = self._load_plugin_metadata(plugin_path=plugin_path, plugin_obj=obj)
                    metadata.star_cls = obj
+                    metadata.config = plugin_config
                    metadata.module = module
                    metadata.root_dir_name = root_dir_name
                    metadata.reserved = reserved
@@ -221,7 +256,7 @@ class PluginManager:
                if metadata.module_path in inactivated_plugins:
                    metadata.activated = False
                    
-                # 执行 initialize 函数
+                # 执行 initialize() 方法
                if hasattr(metadata.star_cls, "initialize"):
                    await metadata.star_cls.initialize()
                    
@@ -292,13 +327,14 @@ class PluginManager:
        if plugin.module_path not in inactivated_plugins:
            inactivated_plugins.append(plugin.module_path)
            
-        inactivated_llm_tools: list = sp.get("inactivated_llm_tools", [])
+        inactivated_llm_tools: list = list(set(sp.get("inactivated_llm_tools", []))) # 后向兼容
        
        # 禁用插件启用的 llm_tool
        for func_tool in llm_tools.func_list:
            if func_tool.handler_module_path == plugin.module_path:
                func_tool.active = False
-                inactivated_llm_tools.append(func_tool.name)
+                if func_tool.name not in inactivated_llm_tools:
+                    inactivated_llm_tools.append(func_tool.name)
    
        sp.put("inactivated_plugins", inactivated_plugins)
        sp.put("inactivated_llm_tools", inactivated_llm_tools)
@@ -323,8 +359,9 @@ class PluginManager:
        plugin.activated = True
        
        
-    def install_plugin_from_file(self, zip_file_path: str):
-        desti_dir = os.path.join(self.plugin_store_path, os.path.basename(zip_file_path))
+    async def install_plugin_from_file(self, zip_file_path: str):
+        dir_name = os.path.basename(zip_file_path).replace(".zip", "")
+        desti_dir = os.path.join(self.plugin_store_path, dir_name)
        self.updator.unzip_file(zip_file_path, desti_dir)

        # remove the zip
@@ -332,6 +369,4 @@ class PluginManager:
            os.remove(zip_file_path)
        except BaseException as e:
            logger.warning(f"删除插件压缩包失败: {str(e)}")
-        
-        self._check_plugin_dept_update()
-
+        await self.reload()
@@ -109,7 +109,7 @@ async def download_file(url: str, path: str, show_progress: bool = False):
    '''
    try:
        async with aiohttp.ClientSession(trust_env=True) as session:
-            async with session.get(url, timeout=120) as resp:
+            async with session.get(url, timeout=1800) as resp:
                if resp.status != 200:
                    raise Exception(f"下载文件失败: {resp.status}")
                total_size = int(resp.headers.get('content-length', 0))
@@ -39,7 +39,6 @@ class RepoZipUpdator():
            else:
                ret = self.github_api_release_parser(result)
        except BaseException:
-            logger.error("解析版本信息失败")
            raise Exception("解析版本信息失败")
        return ret
    
@@ -1,14 +1,12 @@
-import os
-import json
-import traceback
 from .route import Route, Response, RouteContext
 from quart import request
 from astrbot.core.config.default import CONFIG_METADATA_2, DEFAULT_VALUE_MAP
 from astrbot.core.config.astrbot_config import AstrBotConfig
-from astrbot.core.star.config import update_config
 from astrbot.core.core_lifecycle import AstrBotCoreLifecycle
 from astrbot.core.platform.register import platform_registry
 from astrbot.core.provider.register import provider_registry
+from astrbot.core.star.star import star_registry
+from astrbot.core import logger

 def try_cast(value: str, type_: str):
    if type_ == "int" and value.isdigit():
@@ -19,9 +17,9 @@ def try_cast(value: str, type_: str):
    elif type_ == "float" and isinstance(value, int):
        return float(value)

-def validate_config(data, config: AstrBotConfig):
+def validate_config(data, schema: dict, is_core: bool):
    errors = []
-    def validate(data, metadata=CONFIG_METADATA_2, path=""):
+    def validate(data, metadata=schema, path=""):
        for key, meta in metadata.items():
            if key not in data:
                continue
@@ -56,35 +54,33 @@ def validate_config(data, config: AstrBotConfig):
            elif meta["type"] == "object" and not isinstance(value, dict):
                errors.append(f"错误的类型 {path}{key}: 期望是 dict, 得到了 {type(value).__name__}")
                validate(value, meta["items"], path=f"{path}{key}.")
-    validate(data)
+
+    if is_core:
+        for key, group in schema.items():
+            group_meta = group.get("metadata")
+            if not group_meta:
+                continue
+            logger.info(f"验证配置: 组 {key} ...")
+            validate(data, group_meta, path=f"{key}.")
+    else:
+        validate(data, schema)
    
    return errors

-def save_astrbot_config(post_config: dict, config: AstrBotConfig):
+def save_config(post_config: dict, config: AstrBotConfig, is_core: bool = False):
    '''验证并保存配置'''
-    errors = validate_config(post_config, config)
+    errors = None
+    try:
+        if is_core:
+            errors = validate_config(post_config, CONFIG_METADATA_2, is_core)
+        else:
+            errors = validate_config(post_config, config.schema, is_core)
+    except BaseException as e:
+        logger.warning(f"验证配置时出现异常: {e}")
    if errors:
        raise ValueError(f"格式校验未通过: {errors}")
    config.save_config(post_config)
    
-def save_extension_config(post_config: dict):
-    if 'namespace' not in post_config:
-        raise ValueError("Missing key: namespace")
-    if 'config' not in post_config:
-        raise ValueError("Missing key: config")
-
-    namespace = post_config['namespace']
-    config: list = post_config['config'][0]['body']
-    for item in config:
-        key = item['path']
-        value = item['value']
-        typ = item['val_type']
-        if typ == 'int':
-            if not value.isdigit():
-                raise ValueError(f"错误的类型 {namespace}.{key}: 期望是 int, 得到了 {type(value).__name__}")
-            value = int(value)
-        update_config(namespace, key, value)
-
 class ConfigRoute(Route):
    def __init__(self, context: RouteContext, core_lifecycle: AstrBotCoreLifecycle) -> None:
        super().__init__(context)
@@ -92,17 +88,17 @@ class ConfigRoute(Route):
        self.routes = {
            '/config/get': ('GET', self.get_configs),
            '/config/astrbot/update': ('POST', self.post_astrbot_configs),
-            '/config/plugin/update': ('POST', self.post_extension_configs),
+            '/config/plugin/update': ('POST', self.post_plugin_configs),
        }
        self.register_routes()

    async def get_configs(self):
-        # namespace 为空时返回 AstrBot 配置
-        # 否则返回指定 namespace 的插件配置
-        namespace = "" if "namespace" not in request.args else request.args["namespace"]
-        if not namespace:
+        # plugin_name 为空时返回 AstrBot 配置
+        # 否则返回指定 plugin_name 的插件配置
+        plugin_name = request.args.get("plugin_name", None)
+        if not plugin_name:
            return Response().ok(await self._get_astrbot_config()).__dict__
-        return Response().ok(await self._get_extension_config(namespace)).__dict__
+        return Response().ok(await self._get_plugin_config(plugin_name)).__dict__

    async def post_astrbot_configs(self):
        post_configs = await request.json
@@ -110,14 +106,15 @@ class ConfigRoute(Route):
            await self._save_astrbot_configs(post_configs)
            return Response().ok(None, "保存成功~ 机器人正在重载配置。").__dict__
        except Exception as e:
-            traceback.print_exc()
+            logger.error(e)
            return Response().error(str(e)).__dict__
    
-    async def post_extension_configs(self):
+    async def post_plugin_configs(self):
        post_configs = await request.json
+        plugin_name = request.args.get("plugin_name", "unknown")
        try:
-            await self._save_extension_configs(post_configs)
-            return Response().ok(None, "保存成功~ 机器人正在重载配置。").__dict__
+            await self._save_plugin_configs(post_configs, plugin_name)
+            return Response().ok(None, f"保存插件 {plugin_name} 成功~ 机器人正在重载配置。").__dict__
        except Exception as e:
            return Response().error(str(e)).__dict__
            
@@ -141,28 +138,48 @@ class ConfigRoute(Route):
            "config": config
        }

-    async def _get_extension_config(self, namespace: str):
-        path = f"data/config/{namespace}.json"
-        if not os.path.exists(path):
-            return []
-        with open(path, "r", encoding="utf-8-sig") as f:
-            return [{
-                "config_type": "group",
-                "name": namespace + " 插件配置",
-                "description": "",
-                "body": list(json.load(f).values())
-            },]
-
+    async def _get_plugin_config(self, plugin_name: str):
+        ret = {
+            "metadata": None,
+            "config": None
+        }
+        
+        for plugin_md in star_registry:
+            if plugin_md.name == plugin_name:
+                if not plugin_md.config:
+                    break
+                ret['config'] = plugin_md.config # 这是自定义的 Dict 类（AstrBotConfig）
+                ret['metadata'] = {
+                    plugin_name: {
+                        "description": f"{plugin_name} 配置",
+                        "type": "object",
+                        "items": plugin_md.config.schema # 初始化时通过 __setattr__ 存入了 schema
+                    }
+                }
+                break
+            
+        return ret
+        
    async def _save_astrbot_configs(self, post_configs: dict):
        try:
-            save_astrbot_config(post_configs, self.config)
+            save_config(post_configs, self.config, is_core=True)
            self.core_lifecycle.restart()
        except Exception as e:
            raise e
-    
-    async def _save_extension_configs(self, post_configs: dict):
+        
+    async def _save_plugin_configs(self, post_configs: dict, plugin_name: str):
+        md = None
+        for plugin_md in star_registry:
+            if plugin_md.name == plugin_name:
+                md = plugin_md
+        
+        if not md:
+            raise ValueError(f"插件 {plugin_name} 不存在")
+        if not md.config:
+            raise ValueError(f"插件 {plugin_name} 没有注册配置")
+        
        try:
-            save_extension_config(post_configs)
+            save_config(post_configs, md.config)
            self.core_lifecycle.restart()
        except Exception as e:
            raise e
@@ -56,6 +56,7 @@ class PluginRoute(Route):
        try:
            logger.info(f"正在安装插件 {repo_url}")
            await self.plugin_manager.install_plugin(repo_url)
+            self.core_lifecycle.restart()
            logger.info(f"安装插件 {repo_url} 成功。")
            return Response().ok(None, "安装成功。").__dict__
        except Exception as e:
@@ -67,9 +68,10 @@ class PluginRoute(Route):
            file = await request.files
            file = file['file']
            logger.info(f"正在安装用户上传的插件 {file.filename}")
-            file_path = f"data/temp/{uuid.uuid4()}.zip"
+            file_path = f"data/temp/{file.filename}"
            await file.save(file_path)
-            self.plugin_manager.install_plugin_from_file(file_path)
+            await self.plugin_manager.install_plugin_from_file(file_path)
+            self.core_lifecycle.restart()
            logger.info(f"安装插件 {file.filename} 成功")
            return Response().ok(None, "安装成功。").__dict__
        except Exception as e:
@@ -94,6 +96,7 @@ class PluginRoute(Route):
        try:
            logger.info(f"正在更新插件 {plugin_name}")
            await self.plugin_manager.update_plugin(plugin_name)
+            self.core_lifecycle.restart()
            logger.info(f"更新插件 {plugin_name} 成功。")
            return Response().ok(None, "更新成功。").__dict__
        except Exception as e:
@@ -6,6 +6,10 @@ class StaticFileRoute(Route):
        index_ = ['/', '/auth/login', '/config', '/logs', '/extension', '/dashboard/default', '/project-atri', '/console', '/chat']
        for i in index_:
            self.app.add_url_rule(i, view_func=self.index)
+            
+        @self.app.errorhandler(404)
+        async def page_not_found(e):
+            return "404 Not found。如果你初次使用打开面板发现 404，请参考文档: https://astrbot.app/deploy/dashboard-404.html"
    
    async def index(self):
        return await self.app.send_static_file('index.html')
@@ -1,4 +1,5 @@
 import traceback
+import aiohttp
 from .route import Route, Response, RouteContext
 from quart import request
 from astrbot.core.core_lifecycle import AstrBotCoreLifecycle
@@ -43,7 +44,7 @@ class UpdateRoute(Route):
                    }
                ).__dict__
        except Exception as e:
-            logger.error(traceback.format_exc())
+            logger.warning(f"检查更新失败: {str(e)} (不影响除项目更新外的正常使用)")
            return Response().error(e.__str__()).__dict__
    
    async def update_project(self):
@@ -0,0 +1,4 @@
+# What's Changed
+
+- 修复 astrbot_updator 属性缺失与stt_enabled 未初始化 #252
+- 支持消息分段回复
@@ -0,0 +1,8 @@
+# What's Changed
+
+- 修复: TTS 问题
+- 新增: **支持记录非唤醒状态下群聊历史记录(beta)**
+- 优化: 自动删除 deepseek-r1 模型自带的 think 标签
+- 优化: 自动移除 ollama 不支持 tool 的模型的 tool 请求
+- 优化: /t2i 即时生效
+- 优化: gewechat 消息下发异常处理
@@ -0,0 +1,9 @@
+# What's Changed
+
+- 修复: 配置 Validator 不起效的问题
+- 修复: DeepSeek-R1 思考标签问题
+- 修复: 分段回复间隔时间不生效
+- 修复: 修复白名单为空时依然终止事件 #259
+- 修复: 群聊增强某些参数的类型转换问题
+- 新增: 插件支持注册配置，详见 [注册插件配置](https://astrbot.app/dev/plugin.html#%E6%B3%A8%E5%86%8C%E6%8F%92%E4%BB%B6%E9%85%8D%E7%BD%AE-beta)
+- 优化: 插件的禁用/启用逻辑以及函数工具的禁用/启用逻辑
@@ -0,0 +1,6 @@
+# What's Changed
+
+- [gewechat] [修复每次启动astrbot都需要扫码的问题](https://github.com/Soulter/AstrBot/commit/fd5d7dd37a6d74f81a148bbebef8516aa0cb5540)
+- [core] [Provider 重复时不直接报错闪退](https://github.com/Soulter/AstrBot/commit/b61f9be18db9a6b8b3c5b6b36553f66dd2b79375) https://github.com/Soulter/AstrBot/issues/265
+- [core] [弱化更新报错](https://github.com/Soulter/AstrBot/commit/0ba0150fd8ff2062dbe83889163888ba3e33bd49) https://github.com/Soulter/AstrBot/issues/267 
+- 修复 webui 无法从本地上传插件的问题
@@ -0,0 +1,11 @@
+# What's Changed
+
+- [beta] 支持群聊内基于概率的主动回复
+- openai tts 更换模型 #300
+- 增加模型响应后的插件钩子
+- 修复 相同type的provider共享了记忆
+- 优化 人格情景在发现格式不对时仍然加载而不是跳过 #282
+- 修复 Gemini函数调用时，parameters为空对象导致的错误 by @Camreishi
+- 修复 弹出记录报错的问题 #272
+- 优化 移除默认人格
+- 优化 未启用模型提供商时的异常处理
@@ -0,0 +1,12 @@
+# What's Changed
+
+- fix: 修复主动概率回复关闭后仍然回复的问题 #317
+- fix: 尝试修复 gewechat 群聊收不到 at 的回复 #294
+- perf: 移除了默认人格
+- fix: 修复HTTP代理删除后不生效 #319
+- fix: 调用Gemini API输出多余空行问题 #318
+- feat: 添加硅基流动模版
+- fix: 硅基流动 not a vlm 和 tool calling not supported 报错 #305 #291
+- perf: 回复时艾特发送者之后添加空格或换行 #312
+- fix: docker容器内时区不对导致 reminder 时间错误
+- perf: siliconcloud 不支持 tool 的模型
@@ -4,6 +4,8 @@ services:
  astrbot:
    image: soulter/astrbot:latest
    container_name: astrbot
-    network_mode: "host"
+    ports: 
+      - "6180-6200:6180-6200"
+      - "11451:11451"
    volumes:
-      - ./data:/AstrBot/data
+      - ./data:/AstrBot/data
@@ -1,41 +0,0 @@
-<script setup>
-import UiParentCard from '@/components/shared/UiParentCard.vue';
-
-const props = defineProps({
-  config: Array
-});
-</script>
-
-<template>
-  <a v-show="config.length === 0">该插件没有配置</a>
-  <UiParentCard v-for="group in config" :key="group.name" :title="group.name" style="margin-bottom: 16px;">
-    <template v-for="item in group.body">
-      <template v-if="item.config_type === 'item'">
-        <template v-if="item.val_type === 'bool'">
-          <v-switch v-model="item.value" :label="item.name" :hint="item.description" color="primary" inset></v-switch>
-        </template>
-        <template v-else-if="item.val_type === 'str'">
-          <v-text-field v-model="item.value" :label="item.name" :hint="item.description" style="margin-bottom: 8px;"
-            variant="outlined"></v-text-field>
-        </template>
-        <template v-else-if="item.val_type === 'int'">
-          <v-text-field v-model="item.value" :label="item.name" :hint="item.description" style="margin-bottom: 8px;"
-            variant="outlined"></v-text-field>
-        </template>
-        <template v-else-if="item.val_type === 'list'">
-          <span>{{ item.name }}</span>
-          <v-combobox v-model="item.value" chips clearable label="请添加" multiple prepend-icon="mdi-tag-multiple-outline">
-            <template v-slot:selection="{ attrs, item, select, selected }">
-              <v-chip v-bind="attrs" :model-value="selected" closable @click="select" @click:close="remove(item)">
-                <strong>{{ item }}</strong>
-              </v-chip>
-            </template>
-          </v-combobox>
-        </template>
-      </template>
-      <template v-else-if="item.config_type === 'divider'">
-        <v-divider style="margin-top: 8px; margin-bottom: 8px;"></v-divider>
-      </template>
-    </template>
-  </UiParentCard>
-</template>
@@ -1,7 +1,7 @@
 <script setup>
 import ExtensionCard from '@/components/shared/ExtensionCard.vue';
-import ConfigDetailCard from '@/components/shared/ConfigDetailCard.vue';
 import WaitingForRestart from '@/components/shared/WaitingForRestart.vue';
+import AstrBotConfig from '@/components/shared/AstrBotConfig.vue';
 import ConsoleDisplayer from '@/components/shared/ConsoleDisplayer.vue';
 import axios from 'axios';

@@ -52,11 +52,17 @@ import axios from 'axios';
          <v-btn v-else variant="plain" disabled>已安装</v-btn>
        </div>
      </ExtensionCard>
+      
+    </v-col>
+
+    <v-col style="margin-bottom: 16px;" cols="12" md="12">
+      <small ><a href="https://astrbot.app/dev/plugin.html">插件开发文档</a></small> |
+      <small> <a href="https://github.com/Soulter/AstrBot_Plugins_Collection">提交插件仓库</a></small>
    </v-col>

  </v-row>

-  <v-dialog v-model="configDialog" width="750">
+  <v-dialog v-model="configDialog" width="1000">
    <template v-slot:activator="{ props }">
    </template>
    <v-card>
@@ -65,7 +71,8 @@ import axios from 'axios';
      </v-card-title>
      <v-card-text>
        <v-container>
-          <ConfigDetailCard :config="extension_config"></ConfigDetailCard>
+          <AstrBotConfig v-if="extension_config.metadata" :metadata="extension_config.metadata" :iterable="extension_config.config" :metadataKey=curr_namespace></AstrBotConfig>
+          <p v-else>这个插件没有配置</p>
        </v-container>
      </v-card-text>
      <v-card-actions>
@@ -172,9 +179,9 @@ export default {
  name: 'ExtensionPage',
  components: {
    ExtensionCard,
-    ConfigDetailCard,
    WaitingForRestart,
-    ConsoleDisplayer
+    ConsoleDisplayer,
+    AstrBotConfig
  },
  data() {
    return {
@@ -189,7 +196,10 @@ export default {
      snack_success: "success",
      loading_: false,
      configDialog: false,
-      extension_config: {},
+      extension_config: {
+        "metadata": {},
+        "config": {}
+      },
      upload_file: null,
      pluginMarketData: {},
      loadingDialog: {
@@ -364,7 +374,7 @@ export default {
    openExtensionConfig(extension_name) {
      this.curr_namespace = extension_name;
      this.configDialog = true;
-      axios.get('/api/config/get?namespace=' + extension_name).then((res) => {
+      axios.get('/api/config/get?plugin_name=' + extension_name).then((res) => {
        this.extension_config = res.data.data;
        console.log(this.extension_config);
      }).catch((err) => {
@@ -372,10 +382,7 @@ export default {
      });
    },
    updateConfig() {
-      axios.post('/api/config/plugin/update', {
-        "config": this.extension_config,
-        "namespace": this.curr_namespace
-      }).then((res) => {
+      axios.post('/api/config/plugin/update?plugin_name='+this.curr_namespace, this.extension_config.config).then((res) => {
        if (res.data.status === "ok") {
          this.toast(res.data.message, "success");
          this.$refs.wfr.check();
@@ -0,0 +1,128 @@
+import datetime
+import uuid
+import random
+import astrbot.api.star as star
+from astrbot.api.event import AstrMessageEvent
+from astrbot.api.platform import MessageType
+from astrbot.api.provider import ProviderRequest
+from astrbot.api.message_components import Plain, Image
+from astrbot import logger
+from collections import defaultdict
+
+'''
+聊天记忆增强
+'''
+class LongTermMemory:
+    def __init__(self, config: dict, context: star.Context):
+        self.config = config
+        self.context = context
+        self.session_chats = defaultdict(list)
+        """记录群成员的群聊记录"""
+        try:
+            self.max_cnt = int(self.config["group_message_max_cnt"])
+        except BaseException as e:
+            logger.error(e)
+            self.max_cnt = 300
+        self.image_caption = self.config["image_caption"]
+        self.image_caption_prompt = self.config["image_caption_prompt"]
+        
+        self.active_reply = self.config["active_reply"]
+        self.enable_active_reply = self.active_reply.get("enable", False)
+        self.ar_method = self.active_reply["method"]
+        self.ar_possibility = self.active_reply["possibility_reply"]
+        self.ar_prompt = self.active_reply.get("prompt", "")
+        
+        self.put_history_to_prompt = self.config["put_history_to_prompt"]
+        
+    async def remove_session(self, event: AstrMessageEvent) -> int:
+        cnt = 0
+        if event.unified_msg_origin in self.session_chats:
+            cnt = len(self.session_chats[event.unified_msg_origin])
+            del self.session_chats[event.unified_msg_origin]
+        return cnt
+
+    async def get_image_caption(self, image_url: str) -> str:
+        provider = self.context.get_using_provider()
+        response = await provider.text_chat(
+            prompt=self.image_caption_prompt,
+            session_id=uuid.uuid4().hex,
+            image_urls=[image_url],
+            persist=False,
+        )
+        return response.completion_text
+    
+    async def need_active_reply(self, event: AstrMessageEvent) -> bool:
+        if not self.enable_active_reply:
+            return False
+        if event.get_message_type() != MessageType.GROUP_MESSAGE:
+            return False
+        
+        if event.is_at_or_wake_command:
+            # if the message is a command, let it pass
+            return False
+
+        match self.ar_method:
+            case "possibility_reply":
+                trig = random.random() < self.ar_possibility
+                return trig
+            
+        return False
+
+
+    async def handle_message(self, event: AstrMessageEvent):
+        '''仅支持群聊'''
+        if event.get_message_type() == MessageType.GROUP_MESSAGE:
+            datetime_str = datetime.datetime.now().strftime("%H:%M:%S")
+
+            final_message = f"[{event.message_obj.sender.nickname}/{datetime_str}]: "
+
+            for comp in event.get_messages():
+                if isinstance(comp, Plain):
+                    final_message += f" {comp.text}"
+                elif isinstance(comp, Image):
+                    # image_urls.append(comp.url if comp.url else comp.file)
+                    if self.image_caption:
+                        try:
+                            caption = await self.get_image_caption(
+                                comp.url if comp.url else comp.file
+                            )
+                            final_message += f" [Image: {caption}]"
+                        except Exception as e:
+                            logger.error(f"获取图片描述失败: {e}")
+                    else:
+                        final_message += " [Image]"
+            logger.debug(f"ltm | {event.unified_msg_origin} | {final_message}")
+            self.session_chats[event.unified_msg_origin].append(final_message)
+            if len(self.session_chats[event.unified_msg_origin]) > self.max_cnt:
+                self.session_chats[event.unified_msg_origin].pop(0)
+
+    async def on_req_llm(self, event: AstrMessageEvent, req: ProviderRequest):
+        '''当触发 LLM 请求前，调用此方法修改 req'''
+        if event.unified_msg_origin not in self.session_chats:
+            return
+        
+        chats_str = '\n---\n'.join(self.session_chats[event.unified_msg_origin])
+        
+        if self.put_history_to_prompt:
+            prompt = req.prompt
+            req.prompt = f"You are now in a chatroom. The chat history is as follows:\n{chats_str}"
+            req.prompt += f"\nNow, a new message is coming: `{prompt}`. Please react to it. Only output your response and do not output any other information."
+            req.contexts = [] # 清空上下文，当使用了群聊增强，所有聊天记录都在一个prompt中。
+        else:
+            req.system_prompt += "You are now in a chatroom. The chat history is as follows: \n"
+            req.system_prompt += chats_str
+            if self.image_caption:
+                req.system_prompt += (
+                    "The images sent by the members are displayed in text form above."
+                )
+            
+    async def after_req_llm(self, event: AstrMessageEvent):
+        if event.unified_msg_origin not in self.session_chats:
+            return
+
+        if event.get_result() and event.get_result().is_llm_result():
+            final_message = f"[AstrBot/{datetime.datetime.now().strftime('%H:%M:%S')}]: {event.get_result().get_plain_text()}"
+            logger.debug(f"ltm | {event.unified_msg_origin} | {final_message}")
+            self.session_chats[event.unified_msg_origin].append(final_message)
+            if len(self.session_chats[event.unified_msg_origin]) > self.max_cnt:
+                self.session_chats[event.unified_msg_origin].pop(0)
@@ -5,13 +5,17 @@ import astrbot.api.star as star
 import astrbot.api.event.filter as filter
 from astrbot.api.event import AstrMessageEvent, MessageEventResult
 from astrbot.api import sp
-from astrbot.api.provider import Personality, ProviderRequest
+from astrbot.api.provider import Personality, ProviderRequest, LLMResponse
+from astrbot.api.platform import MessageType
 from astrbot.core.utils.io import download_dashboard, get_dashboard_version
 from astrbot.core.config.default import VERSION
+from collections import defaultdict
+from .long_term_memory import LongTermMemory
+from astrbot.core import logger

 from typing import Union

-@star.register(name="astrbot", desc="AstrBot 基础指令集合", author="Soulter", version="4.0.0")
+@star.register(name="astrbot", desc="AstrBot 基础指令结合 + 拓展功能", author="Soulter", version="4.0.0")
 class Main(star.Star):
    def __init__(self, context: star.Context) -> None:
        self.context = context
@@ -20,7 +24,12 @@ class Main(star.Star):
        self.identifier = cfg['provider_settings']['identifier']
        self.enable_datetime = cfg['provider_settings']["datetime_system_prompt"]
        
-        self.kdb_enabled = False
+        self.ltm = None
+        if self.context.get_config()['provider_ltm_settings']['group_icl_enable'] or self.context.get_config()['provider_ltm_settings']['active_reply']['enable']:
+            try:
+                self.ltm = LongTermMemory(self.context.get_config()['provider_ltm_settings'], self.context)
+            except BaseException as e:
+                logger.error(f"聊天增强 err: {e}")
    
    async def _query_astrbot_notice(self):
        try:
@@ -82,7 +91,7 @@ class Main(star.Star):
            active = " (启用)" if tool.active else "(停用)"
            msg += f"- {tool.name}: {tool.description} {active}\n"
            
-        msg += "\n使用 /tool on/off <工具名> 激活或者停用工具。"
+        msg += "\n使用 /tool on/off <工具名> 激活或者停用函数工具。/tool off_all 停用所有函数工具。"
        event.set_result(MessageEventResult().message(msg).use_t2i(False))
        
    @tool.command("on")
@@ -98,6 +107,13 @@ class Main(star.Star):
            event.set_result(MessageEventResult().message(f"停用工具 {tool_name} 成功。"))
        else:
            event.set_result(MessageEventResult().message(f"停用工具 {tool_name} 失败，未找到此工具。"))
+            
+    @tool.command("off_all")
+    async def tool_all_off(self, event: AstrMessageEvent):
+        tm = self.context.get_llm_tool_manager()
+        for tool in tm.func_list:
+            self.context.deactivate_llm_tool(tool.name)
+        event.set_result(MessageEventResult().message(f"停用所有工具成功。"))

    @filter.command("plugin")
    async def plugin(self, event: AstrMessageEvent, oper1: str = None, oper2: str = None):
@@ -194,6 +210,10 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
    async def provider(self, event: AstrMessageEvent, idx: int = None):
        '''查看或者切换 LLM Provider'''
        
+        if not self.context.get_using_provider():
+            event.set_result(MessageEventResult().message("未找到任何 LLM 提供商。请先配置。"))
+            return
+        
        if idx is None:
            ret = "## 当前载入的 LLM 提供商\n"
            for idx, llm in enumerate(self.context.get_all_providers()):
@@ -218,11 +238,27 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo

    @filter.command("reset")
    async def reset(self, message: AstrMessageEvent):
+        
+        if not self.context.get_using_provider():
+            message.set_result(MessageEventResult().message("未找到任何 LLM 提供商。请先配置。"))
+            return
+        
        await self.context.get_using_provider().forget(message.session_id)
-        message.set_result(MessageEventResult().message("重置成功"))
+        ret = "清除会话 LLM 聊天历史成功。"
+        if self.ltm:
+            cnt = await self.ltm.remove_session(event=message)
+            ret += f"\n聊天增强: 已清除 {cnt} 条聊天记录。"
+        
+        message.set_result(MessageEventResult().message(ret))

    @filter.command("model")
    async def model_ls(self, message: AstrMessageEvent, idx_or_name: Union[int, str] = None):
+                
+        if not self.context.get_using_provider():
+            message.set_result(MessageEventResult().message("未找到任何 LLM 提供商。请先配置。"))
+            return
+        
+        
        if idx_or_name is None:
            models = []
            try:
@@ -263,6 +299,12 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo

    @filter.command("history")
    async def his(self, message: AstrMessageEvent, page: int = 1):
+        
+                
+        if not self.context.get_using_provider():
+            message.set_result(MessageEventResult().message("未找到任何 LLM 提供商。请先配置。"))
+            return
+        
        size_per_page = 3
        contexts, total_pages = await self.context.get_using_provider().get_human_readable_context(message.session_id, page, size_per_page)

@@ -282,6 +324,10 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
    @filter.permission_type(filter.PermissionType.ADMIN)
    @filter.command("key")
    async def key(self, message: AstrMessageEvent, index: int=None):
+                
+        if not self.context.get_using_provider():
+            message.set_result(MessageEventResult().message("未找到任何 LLM 提供商。请先配置。"))
+            return
        
        if index is None:
            keys_data = self.context.get_using_provider().get_keys()
@@ -310,6 +356,12 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo

    @filter.command("persona")
    async def persona(self, message: AstrMessageEvent):
+                
+        if not self.context.get_using_provider():
+            message.set_result(MessageEventResult().message("未找到任何 LLM 提供商。请先配置。"))
+            return
+        
+        
        l = message.message_str.split(" ")
        
        curr_persona_name = "无"
@@ -323,6 +375,7 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
 - 设置人格情景: `/persona 人格名`, 如 /persona 编剧
 - 人格情景列表: `/persona list`
 - 人格情景详细信息: `/persona view 人格名`
+- 取消人格: `/persona unset`

 当前人格情景: {curr_persona_name}

@@ -348,6 +401,9 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
            else:
                msg = f"人格{ps}不存在"
            message.set_result(MessageEventResult().message(msg))
+        elif l[1] == "unset":
+            self.context.get_using_provider().curr_personality = None
+            message.set_result(MessageEventResult().message("取消人格成功。"))
        else:
            ps = "".join(l[1:]).strip()
            if persona := next(builtins.filter(
@@ -355,9 +411,9 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
                self.context.provider_manager.personas
            ), None):
                self.context.get_using_provider().curr_personality = persona
-                message.set_result(MessageEventResult().message(f"设置成功。如果您正在切换到不同的人格，请注意使用 /reset 来清空上下文，防止原人格对话影响现人格。"))
+                message.set_result(MessageEventResult().message("设置成功。如果您正在切换到不同的人格，请注意使用 /reset 来清空上下文，防止原人格对话影响现人格。"))
            else:
-                message.set_result(MessageEventResult().message(f"不存在该人格情景。使用 /persona list 查看所有。"))
+                message.set_result(MessageEventResult().message("不存在该人格情景。使用 /persona list 查看所有。"))
    
    @filter.permission_type(filter.PermissionType.ADMIN)
    @filter.command("dashboard_update")
@@ -366,31 +422,6 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
        await download_dashboard()
        yield event.plain_result("管理面板更新完成。")

-    @filter.on_llm_request()
-    async def decorate_llm_req(self, event: AstrMessageEvent, req: ProviderRequest):
-        provider = self.context.get_using_provider()
-        if self.prompt_prefix:
-            req.prompt = self.prompt_prefix + req.prompt
-        if self.identifier:
-            user_id = event.message_obj.sender.user_id
-            user_nickname = event.message_obj.sender.nickname
-            user_info = f"\n[User ID: {user_id}, Nickname: {user_nickname}]\n"
-            req.prompt = user_info + req.prompt
-        if self.enable_datetime:
-            req.system_prompt += f"\nCurrent datetime: {datetime.datetime.now().strftime('%Y-%m-%d %H:%M')}\n"
-        
-        if persona := provider.curr_personality:
-            if prompt := persona['prompt']:
-                req.system_prompt += prompt
-            if mood_dialogs := persona['_mood_imitation_dialogs_processed']:
-                req.system_prompt += "\nHere are few shots of dialogs, you need to imitate the tone of 'B' in the following dialogs to respond:\n"
-                req.system_prompt += mood_dialogs
-            if begin_dialogs := persona["_begin_dialogs_processed"]:
-                req.contexts[:0] = begin_dialogs
-
-        # if provider.curr_personality['prompt']:
-        #     req.system_prompt += f"\n{provider.curr_personality['prompt']}"
-        
    @filter.command("set")
    async def set_variable(self, event: AstrMessageEvent, key: str, value: str):
        session_id = event.get_session_id()
@@ -428,32 +459,111 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
                await platform.logout()
                yield event.plain_result("已登出 gewechat")
                return
+            
+            
+    @filter.platform_adapter_type(filter.PlatformAdapterType.ALL)
+    async def on_message(self, event: AstrMessageEvent):
+        '''群聊记忆增强'''
+        if self.ltm:
+            need_active = await self.ltm.need_active_reply(event)
+                        
+            group_icl_enable = self.context.get_config()['provider_ltm_settings']['group_icl_enable']
+            if group_icl_enable:
+                '''记录对话'''
+                try:
+                    await self.ltm.handle_message(event)
+                except BaseException as e:
+                    logger.error(e)
                
-    @filter.command_group("kdb")
-    def kdb(self):
-        pass
-        
-    @kdb.command("on")
-    async def on_kdb(self, event: AstrMessageEvent):
-        self.kdb_enabled = True
-        curr_kdb_name = self.context.provider_manager.curr_kdb_name
-        if not curr_kdb_name:
-            yield event.plain_result("未载入任何知识库")
-        else:
-            yield event.plain_result(f"知识库已打开。当前载入的知识库: {curr_kdb_name}")
-        
-    @kdb.command("off")
-    async def off_kdb(self, event: AstrMessageEvent):
-        self.kdb_enabled = False
-        yield event.plain_result("知识库已关闭")
-        
+            if need_active:
+                '''主动回复'''
+                provider = self.context.get_using_provider()
+                if not provider:
+                    logger.error("未找到任何 LLM 提供商。请先配置。无法主动回复")
+                    return
+                try:
+                    session_provider_context = provider.session_memory.get(event.session_id)
+                    
+                    prompt = self.ltm.ar_prompt
+                    if not prompt:
+                        prompt = event.message_str
+                    
+                    yield event.request_llm(
+                        prompt=prompt,
+                        func_tool_manager=self.context.get_llm_tool_manager(),
+                        session_id=event.session_id,
+                        contexts=session_provider_context if session_provider_context else []
+                    )
+                except BaseException as e:
+                    logger.error(f"主动回复失败: {e}")
+    
+    
    @filter.on_llm_request()
-    async def on_llm_response(self, event: AstrMessageEvent, req: ProviderRequest):
-        curr_kdb_name = self.context.provider_manager.curr_kdb_name
-        if self.kdb_enabled and curr_kdb_name:
-            mgr = self.context.knowledge_db_manager
-            results = await mgr.retrive_records(curr_kdb_name, req.prompt)
-            if results:
-                req.system_prompt += "\nHere are documents that related to user's query: \n"
-                for result in results:
-                    req.system_prompt += f"- {result}\n"
+    async def decorate_llm_req(self, event: AstrMessageEvent, req: ProviderRequest):
+        '''在请求 LLM 前注入人格信息、Identifier、时间等 System Prompt'''
+        provider = self.context.get_using_provider()
+        if self.prompt_prefix:
+            req.prompt = self.prompt_prefix + req.prompt
+            
+        if self.identifier:
+            user_id = event.message_obj.sender.user_id
+            user_nickname = event.message_obj.sender.nickname
+            user_info = f"\n[User ID: {user_id}, Nickname: {user_nickname}]\n"
+            req.prompt = user_info + req.prompt
+            
+        if self.enable_datetime:
+            req.system_prompt += f"\nCurrent datetime: {datetime.datetime.now().strftime('%Y-%m-%d %H:%M')}\n"
+        
+        if persona := provider.curr_personality:
+            if prompt := persona['prompt']:
+                req.system_prompt += prompt
+            if mood_dialogs := persona['_mood_imitation_dialogs_processed']:
+                req.system_prompt += "\nHere are few shots of dialogs, you need to imitate the tone of 'B' in the following dialogs to respond:\n"
+                req.system_prompt += mood_dialogs
+            if begin_dialogs := persona["_begin_dialogs_processed"]:
+                req.contexts[:0] = begin_dialogs
+                
+        if self.ltm:
+            try:
+                await self.ltm.on_req_llm(event, req)
+            except BaseException as e:
+                logger.error(f"ltm: {e}")
+
+    
+    @filter.after_message_sent()
+    async def after_llm_req(self, event: AstrMessageEvent):
+        '''在 LLM 请求后记录对话'''
+        if self.ltm:
+            try:
+                await self.ltm.after_req_llm(event)
+            except BaseException as e:
+                logger.error(f"ltm: {e}")
+
+    # @filter.command_group("kdb")
+    # def kdb(self):
+    #     pass
+        
+    # @kdb.command("on")
+    # async def on_kdb(self, event: AstrMessageEvent):
+    #     self.kdb_enabled = True
+    #     curr_kdb_name = self.context.provider_manager.curr_kdb_name
+    #     if not curr_kdb_name:
+    #         yield event.plain_result("未载入任何知识库")
+    #     else:
+    #         yield event.plain_result(f"知识库已打开。当前载入的知识库: {curr_kdb_name}")
+        
+    # @kdb.command("off")
+    # async def off_kdb(self, event: AstrMessageEvent):
+    #     self.kdb_enabled = False
+    #     yield event.plain_result("知识库已关闭")
+        
+    # @filter.on_llm_request()
+    # async def on_llm_response(self, event: AstrMessageEvent, req: ProviderRequest):
+    #     curr_kdb_name = self.context.provider_manager.curr_kdb_name
+    #     if self.kdb_enabled and curr_kdb_name:
+    #         mgr = self.context.knowledge_db_manager
+    #         results = await mgr.retrive_records(curr_kdb_name, req.prompt)
+    #         if results:
+    #             req.system_prompt += "\nHere are documents that related to user's query: \n"
+    #             for result in results:
+    #                 req.system_prompt += f"- {result}\n"
@@ -8,7 +8,8 @@ from typing import List
 class Google(SearchEngine):
    def __init__(self) -> None:
        super().__init__()
-        self.proxy = os.environ.get("HTTPS_PROXY")
+        self.proxy = os.environ.get("https_proxy")
+        print(f"Google Search using proxy: {self.proxy}")
    
    async def search(self, query: str, num_results: int) -> List[SearchResult]:
        results = []
@@ -22,6 +22,8 @@ class Main(star.Star):
        self.sogo_search = Sogo()
        self.google = Google()
        
+        self.websearch_link = self.context.get_config()['provider_settings'].get('web_search_link', False)
+        
    async def initialize(self):
        websearch = self.context.get_config()['provider_settings']['web_search']
        if websearch:
@@ -109,8 +111,17 @@ class Main(star.Star):
            except BaseException:
                site_result = ""
            site_result = site_result[:700] + "..." if len(site_result) > 700 else site_result
-            ret += f"{idx}. {i.title} \n{i.snippet}\n{site_result}\n\n"
+            
+            header = f"{idx}. {i.title} "
+            
+            if self.websearch_link and i.url:
+                header += i.url
+            
+            ret += f"{header}\n{i.snippet}\n{site_result}\n\n"
            idx += 1
+            
+        if self.websearch_link:
+            ret += "针对问题，请根据上面的结果分点总结，并且在结尾处附上对应内容的参考链接（如有）。"
        
        return ret

@@ -1,5 +1,4 @@
 pydantic~=2.10.3
-vchat
 aiohttp
 openai
 qq-botpy
@@ -1,33 +0,0 @@
-import torch
-from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor, pipeline
-from datasets import load_dataset
-
-
-device = "cuda:0" if torch.cuda.is_available() else "cpu"
-torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
-
-model_id = "openai/whisper-large-v3"
-
-model = AutoModelForSpeechSeq2Seq.from_pretrained(
-    model_id, torch_dtype=torch_dtype, low_cpu_mem_usage=True
-)
-model.to(device)
-
-processor = AutoProcessor.from_pretrained(model_id)
-
-pipe = pipeline(
-    "automatic-speech-recognition",
-    model=model,
-    tokenizer=processor.tokenizer,
-    feature_extractor=processor.feature_extractor,
-    chunk_length_s=30,
-    batch_size=16,  # batch size for inference - set based on your device
-    torch_dtype=torch_dtype,
-    device=device,
-)
-
-dataset = load_dataset("distil-whisper/librispeech_long", "clean", split="validation")
-sample = dataset[0]["audio"]
-
-result = pipe(sample)
-print(result["text"])
@@ -5,6 +5,7 @@ import asyncio
 from astrbot.core.pipeline.scheduler import PipelineScheduler, PipelineContext
 from astrbot.core.star import PluginManager
 from astrbot.core.config.astrbot_config import AstrBotConfig
+from astrbot.core.config.default import CONFIG_METADATA_2
 from astrbot.core.platform.astr_message_event import AstrMessageEvent
 from astrbot.core.platform.astrbot_message import AstrBotMessage, MessageMember, MessageType
 from astrbot.core.message.message_event_result import MessageChain, ResultContentType
Author	SHA1	Message	Date
Soulter	af48975a6b	chore: v3.4.18	2025-02-03 16:14:27 +08:00
Soulter	6441b149ab	fix: 修复主动概率回复关闭后仍然回复的问题 #317	2025-02-03 14:33:53 +08:00
Soulter	f8892881f8	fix: 尝试修复 gewechat 群聊收不到 at 的回复 #294	2025-02-03 14:28:14 +08:00
Soulter	228aec5401	perf: 移除了默认人格	2025-02-03 14:17:45 +08:00
Soulter	68ad48ff55	fix: 修复HTTP代理删除后不生效 #319	2025-02-03 14:11:50 +08:00
Soulter	541ba64032	fix: 调用Gemini API输出多余空行问题 #318	2025-02-03 13:27:56 +08:00
Soulter	2d870b798c	feat: 添加硅基流动模版	2025-02-03 13:24:22 +08:00
Soulter	0f1fe1ab63	fix: 硅基流动 not a vlm 和 tool calling not supported 报错 #305 # 291 perf: 安装和更新插件后全量重启避免奇奇怪怪的bug feat: 支持 /tool off_all 停用所有函数工具	2025-02-03 13:20:49 +08:00
Soulter	73cc86ddb1	perf: 回复时艾特发送者之后添加空格或换行 #312	2025-02-03 12:04:26 +08:00
Soulter	23128f4be2	perf: 主动回复不支持 qq_official 的 hint	2025-02-03 12:00:05 +08:00
Soulter	92200d0e82	fix: docker容器内时区不对	2025-02-03 01:15:09 +08:00
Soulter	d6e8655792	fix: 抱错时首先移除 tool	2025-02-02 23:17:59 +08:00
Soulter	37076d7920	perf: siliconcloud 不支持 tool 的模型	2025-02-02 23:05:36 +08:00
Soulter	78347ec91b	perf: 当人格长度为1时设置默认人格 feat: 支持取消人格	2025-02-02 22:36:50 +08:00
Soulter	9ded102a0a	chore: v3.4.17	2025-02-02 20:39:26 +08:00
Soulter	59b7d8b8cb	chore: clean code	2025-02-02 20:15:57 +08:00
Soulter	f5b97f6762	perf: 优化 404 提示	2025-02-02 19:55:32 +08:00
Soulter	d47da241af	feat: openai tts 更换模型 #300	2025-02-02 19:47:39 +08:00
Soulter	4611ce15eb	feat: [beta] 支持群聊内基于概率的主动回复	2025-02-02 19:23:46 +08:00
Soulter	aa8c56a688	fix: 相同type的provider共享了记忆	2025-02-02 19:13:47 +08:00
Soulter	ef44d4471a	feat: 增加模型响应后的插件钩子 remove: 移除了默认的r1过滤	2025-02-02 16:42:21 +08:00
Soulter	5581eae957	fix: deepseek-r1模型存在遗留“</think>”的问题 #279 Open	2025-02-02 14:59:17 +08:00
Soulter	ec46dfaac9	perf: 人格情景在发现格式不对时仍然加载而不是跳过 #282	2025-02-02 14:59:17 +08:00
Soulter	6042a047bd	修复Gemini函数调用时，parameters为空对象导致的错误 fix: 修复Gemini函数调用时，parameters为空对象导致的错误	2025-02-02 14:50:34 +08:00
Soulter	6ca9e2a753	perf: websearch 可选配置引用链接 #287	2025-02-02 14:42:13 +08:00
Camreishi	618eabfe5c	fix: 修复Gemini函数调用时，parameters为空对象导致的错误 Closes #288	2025-02-02 13:25:08 +08:00
Soulter	bb5db2e9d0	fix: 修复弹出记录报错的问题 #272	2025-02-02 13:24:05 +08:00
Soulter	97e4d169b3	perf: 未启用模型提供商时的异常处理	2025-02-02 11:23:33 +08:00
Soulter	50e44b1473	perf: 移除默认人格	2025-02-02 11:12:17 +08:00
Soulter	38588dd3fa	update compose	2025-02-02 00:11:55 +08:00
Soulter	d183388347	perf: 去除gewechat默认配置	2025-02-01 23:20:25 +08:00
Soulter	1e69d59384	fix: 配置提示 typo	2025-02-01 22:56:34 +08:00
Soulter	00f008f94d	Update compose.yml	2025-02-01 21:31:24 +08:00
Soulter	3c28001a74	v3.4.16	2025-02-01 19:31:59 +08:00
Soulter	76a6218be6	fix: 修复webui无法从本地上传插件的问题	2025-02-01 19:31:29 +08:00
Soulter	6c1de1bbd6	Update README.md	2025-02-01 16:19:01 +08:00
Soulter	d7678081da	perf: Provider 重复时不直接报错闪退 #265	2025-02-01 14:36:41 +08:00
Soulter	5e4ba563cb	perf: 弱化更新报错 #267	2025-02-01 14:29:39 +08:00
Soulter	8afbe77b0a	Update README.md	2025-02-01 12:11:58 +08:00
Soulter	2ef139b59a	fix: 修复每次启动astrbot都需要微信扫码的问题	2025-01-31 01:28:49 +08:00
Soulter	1f0d2d9b89	fix: QQ官方机器人开启 reply with metion 和 reply with quote 后，无法正常回复消息 #244	2025-01-30 01:36:25 +08:00
Soulter	37a1f144ab	chore: update changelog of 3.4.15	2025-01-30 00:32:50 +08:00
Soulter	9a7a654596	perf: 插件处于禁用状态时其所属的函数调用工具不可被启用 #254	2025-01-30 00:27:10 +08:00
Soulter	9abccd63cf	chore: remove stt.py	2025-01-29 23:47:50 +08:00
Soulter	93fea77182	chore: bump to v3.4.15	2025-01-29 23:43:09 +08:00
Soulter	19797243f6	perf: 增加插件链接	2025-01-29 19:56:09 +08:00
Soulter	c9c733d925	Merge branch 'dev'	2025-01-29 19:43:52 +08:00
Soulter	a7d7678c78	fix: 修复白名单为空时依然终止事件 #259	2025-01-29 17:17:27 +08:00
Soulter	c0911921c7	feat: 配置Schema以及插件支持配置	2025-01-29 16:54:57 +08:00
Soulter	4a4241d57a	Update README.md	2025-01-29 13:26:51 +08:00
Soulter	c9426bb6eb	config	2025-01-29 12:25:54 +08:00
Soulter	db4abd169a	fix: 优化分段回复	2025-01-28 14:42:15 +08:00
Soulter	80b6958599	fix: 修复 config validator 不起效的问题	2025-01-28 14:18:21 +08:00
Soulter	80058c781a	fix: 修复r1思考标签问题和分段回复间隔时间问题	2025-01-28 14:03:10 +08:00
Soulter	44bd2e36f3	Update README.md	2025-01-28 02:15:11 +08:00
Soulter	3589a5e5be	perf: 强化ltm异常处理	2025-01-27 21:47:35 +08:00
Soulter	13ef033f0e	fix: 群聊增强的参数类型转换	2025-01-27 21:40:20 +08:00
Soulter	3f8c68bbca	fix: f-string expression part cannot include a backslash long_term_memory.py, line 69	2025-01-27 21:01:50 +08:00
Soulter	4275cea82b	chore: v3.4.14	2025-01-27 20:09:03 +08:00
Soulter	a0bcb5339a	perf: 自动删除 deepseek-r1 模型自带的 think 标签	2025-01-27 20:04:39 +08:00
Soulter	43deec4a4b	Merge pull request #255 from Soulter/feat-ltm 支持记录非唤醒状态下群聊历史记录	2025-01-27 20:02:43 +08:00
Soulter	2bc433a30b	feat: 支持记录非唤醒状态下群聊历史记录	2025-01-27 20:00:32 +08:00
Soulter	eb2b395932	perf: /t2i 即时生效	2025-01-27 19:33:38 +08:00
Soulter	2bfd1c0bf2	perf: 自动移除 ollama 不支持 tool 的模型的 tool 请求	2025-01-27 19:25:28 +08:00
Soulter	7228c4b13f	fix: 修复 TTS 部分变量名错误导致请求失败	2025-01-27 18:45:34 +08:00
Soulter	9351d7471f	perf: 优化 gewechat 消息下发异常处理	2025-01-27 18:11:31 +08:00
Soulter	1cf49998bc	Update README.md	2025-01-27 11:34:27 +08:00
Soulter	6ae86597e8	chore: v3.4.13	2025-01-26 16:51:13 +08:00
Soulter	c578ff25bd	fix: stt_enabled 未初始化 #252	2025-01-26 16:51:02 +08:00
Soulter	2934a3e3be	chore: logo	2025-01-26 15:18:23 +08:00
Soulter	ceaa69da75	feat: 支持消息分段回复	2025-01-26 13:45:32 +08:00
Soulter	fa8e731576	Update README.md	2025-01-25 22:45:47 +08:00