添加关于 LLM 配置的文档

添加基础的 LLM 支持
修复无法更改图片的问题
2025-11-05 20:36:51 +08:00 · 2025-11-05 18:40:13 +08:00 · 2025-11-01 20:59:58 +08:00 · 2025-11-01 18:42:47 +08:00 · 2025-11-01 18:29:10 +08:00 · 2025-11-01 17:52:05 +08:00
34 changed files with 1352 additions and 306 deletions
--- a/.drone.yml
+++ b/.drone.yml
@ -38,6 +38,17 @@ steps:
        path: /var/run/docker.sock
    commands:
      - docker run --rm gitea.service.jazzwhom.top/mttu-developers/konabot:nightly-${DRONE_COMMIT_SHA} python scripts/test_plugin_load.py
  - name: 发送构建结果到 ntfy
    image: parrazam/drone-ntfy
    when:
      status: [success, failure]
    settings:
      url: https://ntfy.service.jazzwhom.top
      topic: drone_ci
      tags:
        - drone-ci
      token:
        from_secret: NTFY_TOKEN
 volumes:
  - name: docker-socket
@ -74,6 +85,17 @@ steps:
    volumes:
      - name: docker-socket
        path: /var/run/docker.sock
  - name: 发送构建结果到 ntfy
    image: parrazam/drone-ntfy
    when:
      status: [success, failure]
    settings:
      url: https://ntfy.service.jazzwhom.top
      topic: drone_ci
      tags:
        - drone-ci
      token:
        from_secret: NTFY_TOKEN
 volumes:
  - name: docker-socket
--- a/.vscode/settings.json
+++ b/.vscode/settings.json
@ -0,0 +1,3 @@
 {
    "python.REPL.enableREPLSmartSend": false
 }
--- a/README.md
+++ b/README.md
@ -1,4 +1,4 @@
-# 此方 bot！
+# konabot
 在 MTTU 内部使用的 bot 一只。
@ -63,12 +63,16 @@ code .
 配置 `ENABLE_CONSOLE=false`
 #### 配置并支持 LLM（大语言模型）
 详见[LLM 配置文档](/docs/LLM.md)。
 ### 运行
 使用命令行手动启动 Bot：
 ```bash
-poetry run watchfiles bot.main konabot
+poetry run watchfiles bot.main . --filter scripts.watch_filter.filter
 ```
 如果你不希望自动重载，只是想运行 Bot，可以直接运行：
--- a/assets/webpage/ac/assets/background.png
+++ b/assets/webpage/ac/assets/background.png
--- a/assets/webpage/ac/index.html
+++ b/assets/webpage/ac/index.html
@ -0,0 +1,76 @@
 <html>
    <head>
        <meta charset="UTF-8">
        <meta name="viewport" content="width=device-width, initial-scale=1.0">
        <title>空调炸炸排行榜</title>
    </head>
    <body>
        <div class="box">
            <div class="text">位居全球第 <span id="ranking" class="ranking">200</span>！</div>
            <div class="text-2">您的群总共坏了 <span id="number" class="number">200</span> 台空调</div>
            <img class="background" src="./assets/background.png" alt="空调炸炸排行榜">
        </div>
    </body>
    <style>
        .box {
            position: relative;
            width: 1024px;
        }
        .number {
            font-size: 2em;
            color: #ffdd00;
            text-shadow: 3px 3px 6px rgba(0, 0, 0, 0.7);
            font-weight: bold;
            font-stretch: 50%;
            max-width: 520px;
            word-wrap: break-word;
            line-height: 0.8em;
        }
        .background {
            width: 1024px;
        }
        .text {
            position: absolute;
            top: 125px;
            width: 100%;
            font-size: 72px;
            color: white;
            text-shadow: 2px 2px 4px rgba(0, 0, 0, 0.7);
            font-weight: bolder;
            display: flex;
            align-items: baseline;
            justify-content: center;
        }
        .text-2 {
            position: absolute;
            top: 50px;
            width: 100%;
            font-size: 48px;
            color: white;
            text-shadow: 2px 2px 4px rgba(0, 0, 0, 0.7);
            font-weight: bolder;
            display: flex;
            align-items: baseline;
            justify-content: center;
        }
        .ranking {
            font-size: 2em;
            color: #ff0000;
            -webkit-text-stroke: #ffffff 2px;
            text-shadow: 3px 3px 6px rgba(0, 0, 0, 0.7);
            font-weight: bold;
            font-stretch: 50%;
        }
    </style>
    <script>
        // 从 URL 参数中获取 number 的值
        const urlParams = new URLSearchParams(window.location.search);
        const number = urlParams.get('number');
        // 将 number 显示在页面上
        document.getElementById('number').textContent = number;
        // 从 URL 参数中获取 ranking 的值
        const ranking = urlParams.get('ranking');
        // 将 ranking 显示在页面上
        document.getElementById('ranking').textContent = ranking;
    </script>
 </html>
--- a/docs/LLM.md
+++ b/docs/LLM.md
@ -0,0 +1,65 @@
 # 大语言模型平台接入
 为实现更多神秘小功能，此方 Bot 需要接入 AI。如果你需要参与开发或测试涉及 AI 的相关功能，麻烦请根据下面的文档继续操作。
 ## 配置项目接入 AI
 AI 相关的配置文件在 `data/config/llm.json` 文件中。示例格式如下，这也将是到时候在云端的配置文件格式（给出的模型都会有）：
 ```json
 {
  "llms": {
    "Qwen2.5-7B-Instruct": {
      "base_url": "https://api.siliconflow.cn/v1",
      "api_key": "sk-XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX",
      "model_name": "Qwen/Qwen2.5-7B-Instruct"
    },
    "qwen3-max": {
      "base_url": "https://dashscope.aliyuncs.com/compatible-mode/v1",
      "api_key": "sk-XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX",
      "model_name": "qwen3-max"
    }
  },
  "default_llm": "Qwen2.5-7B-Instruct"
 }
 ```
 其中，形如 `qwen3-max` 的名称，是你在程序中调用 LLM 使用的键名。若不给出，则会默认使用配置文件中指定的默认模型。
 ```python
 from konabot.common.llm import get_llm
 llm = get_llm()               # 获得的是 Qwen2.5-7B-Instruct 模型
 llm = get_llm("qwen3-max")    # 获得的是 qwen3-max 模型
 message = await llm.chat([
  { "role": "system", "content": "你是一只猫娘" },
  { "role": "user", "content": "晚上好呀！" },
 ], timeout=None, max_tokens=16384)
 # 获得了的是 openai.types.chat.ChatCompletionMessage 对象
 print(f"AI 返回值：{message.content}")    # 注意 content 可能为 None，需要做空值检测
 client = llm.get_openai_client()    # 获得的是一个 OpenAI Client 对象，可以做更多操作
                                    # 例如，调用 Embedding 模型来做知识库向量化等工作
 ```
 ## 本项目使用的模型清单
 为了便利大家使用，我在这里给出该项目将会使用的模型清单，请根据你的开发需求注册并选择你最喜欢的模型。如果需要接入新的模型，或者使用到文档之外的模型，欢迎在这里给出！
 ### 硅基流动 Qwen/Qwen2.5-7B-Instruct
 一个 7B 大小的 AI 模型。其性能不太能指望，但是它小，而且比较快，可以做一些轻量的操作。
 该模型是免费的，但是也需要你注册[硅基流动](https://cloud.siliconflow.cn/me/models)账号，并生成 `api_key` 添加到配置文件中。
 ### 通义千问 qwen3-max
 贵但是很先进的最新模型，其能力可以信赖。但是不要拿它做大量工作哦！
 在[百炼大模型平台](https://bailian.console.aliyun.com/)注册账号并申请 `api_key`，新用户会赠送 1M tokens，足够做测试了。
 ## 安全须知
 请注意提防 AI 越狱等情况。
--- a/konabot/common/data_man.py
+++ b/konabot/common/data_man.py
@ -19,12 +19,12 @@ class DataManager(Generic[T]):
        if not self.fp.exists():
            return self.cls()
        try:
-            return self.cls.model_validate_json(self.fp.read_text())
+            return self.cls.model_validate_json(self.fp.read_text("utf-8"))
        except ValidationError:
            return self.cls()
    def save(self, data: T):
-        self.fp.write_text(data.model_dump_json())
+        self.fp.write_text(data.model_dump_json(), "utf-8")
    @asynccontextmanager
    async def get_data(self):
--- a/konabot/common/llm/init.py
+++ b/konabot/common/llm/init.py
@ -0,0 +1,64 @@
 from typing import Any
 import openai
 from loguru import logger
 from openai.types.chat import ChatCompletion, ChatCompletionMessage, ChatCompletionMessageParam
 from pydantic import BaseModel, Field
 from konabot.common.path import CONFIG_PATH
 LLM_CONFIG_PATH = CONFIG_PATH / 'llm.json'
 if not LLM_CONFIG_PATH.exists():
    LLM_CONFIG_PATH.write_text("{}")
 class LLMInfo(BaseModel):
    base_url: str
    api_key: str
    model_name: str
    def get_openai_client(self):
        return openai.AsyncClient(
            api_key=self.api_key,
            base_url=self.base_url,
        )
    async def chat(
        self,
        messages: list[ChatCompletionMessageParam],
        timeout: float | None = 30.0,
        max_tokens: int | None = None,
        **kwargs: Any,
    ) -> ChatCompletionMessage:
        logger.info(f"调用 LLM: BASE_URL={self.base_url} MODEL_NAME={self.model_name}")
        completion: ChatCompletion = await self.get_openai_client().chat.completions.create(
            messages=messages,
            model=self.model_name,
            max_tokens=max_tokens,
            timeout=timeout,
            stream=False,
            **kwargs,
        )
        choice = completion.choices[0]
        logger.info(
            f"调用 LLM 完成: BASE_URL={self.base_url} MODEL_NAME={self.model_name} REASON={choice.finish_reason}"
        )
        return choice.message
 class LLMConfig(BaseModel):
    llms: dict[str, LLMInfo] = Field(default_factory=dict)
    default_llm: str = "Qwen2.5-7B-Instruct"
 llm_config = LLMConfig.model_validate_json(LLM_CONFIG_PATH.read_text())
 def get_llm(llm_model: str | None = None):
    if llm_model is None:
        llm_model = llm_config.default_llm
    if llm_model not in llm_config.llms:
        raise NotImplementedError("LLM 未配置，该功能无法使用")
    return llm_config.llms[llm_model]
--- a/konabot/common/longtask.py
+++ b/konabot/common/longtask.py
@ -51,6 +51,10 @@ class LongTaskTarget(BaseModel):
    target_id: str
    "沟通对象的 ID"
    @property
    def is_private_chat(self):
        return self.channel_id.startswith(QQ_PRIVATE_CHAT_CHANNEL_PREFIX)
    async def send_message(self, msg: UniMessage | str, at: bool = True) -> bool:
        try:
            bot = nonebot.get_bot(self.self_id)
@ -236,7 +240,7 @@ def handle_long_task(callback_id: str):
 def _load_longtask_data() -> LongTaskModuleData:
    try:
-        txt = LONGTASK_DATA_DIR.read_text()
+        txt = LONGTASK_DATA_DIR.read_text("utf-8")
        return LongTaskModuleData.model_validate_json(txt)
    except (FileNotFoundError, ValidationError) as e:
        logger.info(f"取得 LongTask 数据时出现问题：{e}")
@ -247,7 +251,7 @@ def _load_longtask_data() -> LongTaskModuleData:
 def _save_longtask_data(data: LongTaskModuleData):
-    LONGTASK_DATA_DIR.write_text(data.model_dump_json())
+    LONGTASK_DATA_DIR.write_text(data.model_dump_json(), "utf-8")
@asynccontextmanager
--- a/konabot/common/nb/qq_broadcast.py
+++ b/konabot/common/nb/qq_broadcast.py
@ -1,10 +1,13 @@
 from typing import Any, cast
 import nonebot
 from nonebot_plugin_alconna import UniMessage
 from nonebot.adapters.onebot.v11 import Bot as OBBot
 from nonebot_plugin_alconna import UniMessage
-async def qq_broadcast(groups: list[str], msg: UniMessage[Any]):
+async def qq_broadcast(groups: list[str], msg: UniMessage[Any] | str):
    if isinstance(msg, str):
        msg = UniMessage.text(msg)
    bots: dict[str, OBBot] = {}
    # group_id -> bot_id
--- a/konabot/common/path.py
+++ b/konabot/common/path.py
@ -6,6 +6,7 @@ FONTS_PATH = ASSETS_PATH / "fonts"
 SRC_PATH = Path(__file__).resolve().parent.parent
 DATA_PATH = SRC_PATH.parent / "data"
 LOG_PATH = DATA_PATH / "logs"
 CONFIG_PATH = DATA_PATH / "config"
 DOCS_PATH = SRC_PATH / "docs"
 DOCS_PATH_MAN1 = DOCS_PATH / "user"
@ -19,3 +20,5 @@ if not DATA_PATH.exists():
 if not LOG_PATH.exists():
    LOG_PATH.mkdir()
 CONFIG_PATH.mkdir(exist_ok=True)
--- a/konabot/common/username.py
+++ b/konabot/common/username.py
@ -0,0 +1,54 @@
 import re
 import nonebot
 from nonebot.adapters.onebot.v11 import Bot as OBBot
 class UsernameManager:
    grouped_data: dict[int, dict[int, str]]
    individual_data: dict[int, str]
    def __init__(self) -> None:
        self.grouped_data = {}
        self.individual_data = {}
    async def update(self):
        for bot in nonebot.get_bots().values():
            if isinstance(bot, OBBot):
                for user in await bot.get_friend_list():
                    uid = user["user_id"]
                    nickname = user["nickname"]
                    self.individual_data[uid] = nickname
                for group in await bot.get_group_list():
                    gid = group["group_id"]
                    for member in await bot.get_group_member_list(group_id=gid):
                        uid = member["user_id"]
                        card = member.get("card", "")
                        nickname = member.get("nickname", "")
                        if card:
                            self.grouped_data.setdefault(gid, {})[uid] = card
                        if nickname:
                            self.individual_data[uid] = nickname
    def get(self, qqid: int, groupid: int | None = None) -> str:
        if groupid is not None and groupid in self.grouped_data:
            n = self.grouped_data[groupid].get(qqid)
            if n is not None:
                return n
        if qqid in self.individual_data:
            return self.individual_data[qqid]
        return str(qqid)
 manager = UsernameManager()
 def get_username(qqid: int | str, group: int | str | None = None):
    if isinstance(group, str):
        group = None if not re.match(r"^\d+$", group) else int(group)
    if isinstance(qqid, str):
        if re.match(r"^\d+$", qqid):
            qqid = int(qqid)
        else:
            return qqid
    return manager.get(qqid, group)
--- a/konabot/common/web_render/init.py
+++ b/konabot/common/web_render/init.py
@ -80,6 +80,30 @@ class WebRenderer:
        page = cls.page_pool[page_id]
        return await instance.render_with_page(page, url, target, params=params, other_function=other_function, timeout=timeout)
    @classmethod
    async def render_file(
        cls,
        file_path: str,
        target: str,
        params: dict = {},
        other_function: PageFunction | None = None,
        timeout: int = 30,
    ) -> bytes:
        '''
        访问指定本地文件URL并返回截图
        :param file_path: 目标文件路径
        :param target: 渲染目标，如 ".box"、"#main" 等CSS选择器
        :param timeout: 页面加载超时时间，单位秒
        :param params: URL键值对参数
        :param other_function: 其他自定义操作函数，接受page参数
        :return: 截图的字节数据
        '''
        instance = await cls.get_browser_instance()
        logger.debug(f"Using WebRendererInstance {id(instance)} to render file {file_path} targeting {target}")
        return await instance.render_file(file_path, target, params=params, other_function=other_function, timeout=timeout)
    @classmethod
    async def close_persistent_page(cls, page_id: str) -> None:
        '''
@ -154,6 +178,10 @@ class WebRendererInstance:
        async with self.lock:
            screenshot = await self.inner_render(page, url, target, index, params, other_function, timeout)
            return screenshot
    async def render_file(self, file_path: str, target: str, index: int = 0, params: dict = {}, other_function: callable = None, timeout: int = 30) -> bytes:
        file_path = "file:///" + str(file_path).replace("\\", "/")
        return await self.render(file_path, target, index, params, other_function, timeout)
    async def inner_render(self, page: Page, url: str, target: str, index: int = 0, params: dict = {}, other_function: callable = None, timeout: int = 30) -> bytes:
        logger.debug(f"Navigating to {url} with timeout {timeout}")
--- a/konabot/docs/concepts/中间答案.txt
+++ b/konabot/docs/concepts/中间答案.txt
@ -0,0 +1,11 @@
 关于「中间答案」或者「提示」：
 在 KonaPH 中，当有人发送「提交答案 答案」时，会检查答案是否符合你设置的中间答案的 pattern。这个 pattern 可以有两种方式：
 - 纯文本的完整匹配：你设置的 pattern 如果和提交的答案完全相等，则会触发提示。
 - regex 匹配：你设置的 pattern 如果以斜杠（/）开头和结尾，就会检查提交的答案是否匹配正则表达式。注意 ^ 和 $ 符号的使用。
    - 例如：/^commit$/ 会匹配 commit，不会匹配 acommit、Commit 等。
    - 而如果是 /commit/，则会匹配 commit、acommit，而不会匹配 Commit。
    - 无法使用 Javascript 的字符串声明模式，例如，/case_insensitive/i 就不会被视作一个正则表达式。
 一个提示是提示，还是中间答案，取决于它是否有 checkpoint 标记。如果有 checkpoint 标记，则会提示用户「你回答了一个中间答案」，并且这个中间答案的回答会在排行榜中显示。
--- a/konabot/docs/concepts/占位.md
+++ b/konabot/docs/concepts/占位.md
--- a/konabot/docs/lib/占位.md
+++ b/konabot/docs/lib/占位.md
--- a/konabot/plugins/air_conditioner/init.py
+++ b/konabot/plugins/air_conditioner/init.py
@ -1,13 +1,19 @@
 from io import BytesIO
 from typing import Optional, Union
 import cv2
 from nonebot.adapters import Event as BaseEvent
 from nonebot.adapters.console.event import MessageEvent as ConsoleMessageEvent
 from nonebot.adapters.discord.event import MessageEvent as DiscordMessageEvent
 from nonebot_plugin_alconna import Alconna, AlconnaMatcher, Args, UniMessage, on_alconna
 from PIL import Image
 import numpy as np
 from konabot.common.longtask import DepLongTaskTarget
 from konabot.common.path import ASSETS_PATH
-from konabot.plugins.air_conditioner.ac import AirConditioner, generate_ac_image
+from konabot.common.web_render import WebRenderer
 from konabot.plugins.air_conditioner.ac import AirConditioner, CrashType, generate_ac_image, wiggle_transform
 import random
 import math
 def get_ac(id: str) -> AirConditioner:
    ac = AirConditioner.air_conditioners.get(id)
@ -70,42 +76,58 @@ async def _(event: BaseEvent, target: DepLongTaskTarget):
    await send_ac_image(evt, ac)
 evt = on_alconna(Alconna(
-    "空调升温"
+    "空调升温",
    Args["temp?", Optional[Union[int, float]]]  # 可选参数，升温的度数，默认为1
 ), use_cmd_start=True, use_cmd_sep=False, skip_for_unmatch=True)
@evt.handle()
-async def _(event: BaseEvent, target: DepLongTaskTarget):
+async def _(event: BaseEvent, target: DepLongTaskTarget, temp: Optional[Union[int, float]] = 1):
    if temp <= 0:
        return
    id = target.channel_id
    ac = get_ac(id)
    if not ac.on or ac.burnt == True or ac.frozen == True:
        await send_ac_image(evt, ac)
        return
-    ac.temperature += 1
+    ac.temperature += temp
    if ac.temperature > 40:
-        # 打开爆炸图片
+        # 根据温度随机出是否爆炸，40度开始，呈指数增长
-        with open(ASSETS_PATH / "img" / "other" / "boom.jpg", "rb") as f:
+        possibility = -math.e ** ((40-ac.temperature) / 50) + 1
-            output = BytesIO()
+        if random.random() < possibility:
-            Image.open(f).save(output, format="GIF")
+            # 打开爆炸图片
-            await evt.send(await UniMessage().image(raw=output).export())
+            with open(ASSETS_PATH / "img" / "other" / "boom.jpg", "rb") as f:
-        ac.burnt = True
+                output = BytesIO()
-        await evt.send("太热啦，空调炸了！")
+                # 爆炸抖动
-        return
+                frames = wiggle_transform(np.array(Image.open(f)), intensity=5)
                pil_frames = [Image.fromarray(frame) for frame in frames]
                pil_frames[0].save(output, format="GIF", save_all=True, append_images=pil_frames[1:], loop=0, duration=35, disposal=2)
                output.seek(0)
                await evt.send(await UniMessage().image(raw=output).export())
            ac.broke_ac(CrashType.BURNT)
            await evt.send("太热啦，空调炸了！")
            return
    await send_ac_image(evt, ac)
 evt = on_alconna(Alconna(
-    "空调降温"
+    "空调降温",
    Args["temp?", Optional[Union[int, float]]]  # 可选参数，降温的度数，默认为1
 ), use_cmd_start=True, use_cmd_sep=False, skip_for_unmatch=True)
@evt.handle()
-async def _(event: BaseEvent, target: DepLongTaskTarget):
+async def _(event: BaseEvent, target: DepLongTaskTarget, temp: Optional[Union[int, float]] = 1):
    if temp <= 0:
        return
    id = target.channel_id
    ac = get_ac(id)
    if not ac.on or ac.burnt == True or ac.frozen == True:
        await send_ac_image(evt, ac)
        return
-    ac.temperature -= 1
+    ac.temperature -= temp
    if ac.temperature < 0:
-        ac.frozen = True
+        # 根据温度随机出是否冻结，0度开始，呈指数增长
        possibility = -math.e ** (ac.temperature / 50) + 1
        if random.random() < possibility:
            ac.broke_ac(CrashType.FROZEN)
    await send_ac_image(evt, ac)
 evt = on_alconna(Alconna(
@ -117,4 +139,24 @@ async def _(event: BaseEvent, target: DepLongTaskTarget):
    id = target.channel_id
    ac = get_ac(id)
    ac.change_ac()
-    await send_ac_image(evt, ac)
+    await send_ac_image(evt, ac)
 evt = on_alconna(Alconna(
    "空调炸炸排行榜",
 ), use_cmd_start=True, use_cmd_sep=False, skip_for_unmatch=True)
@evt.handle()
 async def _(event: BaseEvent, target: DepLongTaskTarget):
    id = target.channel_id
    ac = get_ac(id)
    number, ranking = ac.get_crashes_and_ranking()
    params = {
        "number": number,
        "ranking": ranking
    }
    image = await WebRenderer.render_file(
        file_path=ASSETS_PATH / "webpage" / "ac" / "index.html",
        target=".box",
        params=params
    )
    await evt.send(await UniMessage().image(raw=image).export())
--- a/konabot/plugins/air_conditioner/ac.py
+++ b/konabot/plugins/air_conditioner/ac.py
@ -1,3 +1,4 @@
 from enum import Enum
 from io import BytesIO
 import cv2
@ -5,6 +6,12 @@ import numpy as np
 from PIL import Image, ImageDraw, ImageFont
 from konabot.common.path import ASSETS_PATH, FONTS_PATH
 from konabot.common.path import DATA_PATH
 import json
 class CrashType(Enum):
    BURNT = 0
    FROZEN = 1
 class AirConditioner:
    air_conditioners: dict[str, "AirConditioner"] = {}
@ -23,6 +30,60 @@ class AirConditioner:
        self.on = False
        self.temperature = 24  # 重置为默认温度
    def broke_ac(self, crash_type: CrashType):
        '''
        让空调坏掉，并保存数据
        :param crash_type: CrashType 枚举，表示空调坏掉的类型
        '''
        match crash_type:
            case CrashType.BURNT:
                self.burnt = True
            case CrashType.FROZEN:
                self.frozen = True
        self.save_crash_data(crash_type)
    def save_crash_data(self, crash_type: CrashType):
        '''
        如果空调爆炸了，就往本地的 ac_crash_data.json 里该 id 的记录加一
        '''
        data_file = DATA_PATH / "ac_crash_data.json"
        crash_data = {}
        if data_file.exists():
            with open(data_file, "r", encoding="utf-8") as f:
                crash_data = json.load(f)
        if self.id not in crash_data:
            crash_data[self.id] = {"burnt": 0, "frozen": 0}
        match crash_type:
            case CrashType.BURNT:
                crash_data[self.id]["burnt"] += 1
            case CrashType.FROZEN:
                crash_data[self.id]["frozen"] += 1
        with open(data_file, "w", encoding="utf-8") as f:
            json.dump(crash_data, f, ensure_ascii=False, indent=4)
    def get_crashes_and_ranking(self) -> tuple[int, int]:
        '''
        获取该群在全国空调损坏的数量与排行榜的位置
        '''
        data_file = DATA_PATH / "ac_crash_data.json"
        if not data_file.exists():
            return 0, 1
        with open(data_file, "r", encoding="utf-8") as f:
            crash_data = json.load(f)
        ranking_list = []
        for gid, record in crash_data.items():
            total = record.get("burnt", 0) + record.get("frozen", 0)
            ranking_list.append((gid, total))
        ranking_list.sort(key=lambda x: x[1], reverse=True)
        total_crashes = crash_data.get(self.id, {}).get("burnt", 0) + crash_data.get(self.id, {}).get("frozen", 0)
        rank = 1
        for gid, total in ranking_list:
            if gid == self.id:
                break
            rank += 1
        return total_crashes, rank
 def text_to_transparent_image(text, font_size=40, padding=0, text_color=(0, 0, 0)):
    """
    将文本转换为带透明背景的图像，图像大小刚好包含文本
@ -168,13 +229,13 @@ def precise_blend_with_perspective(background, foreground, corners):
    return result
-def wiggle_transform(image) -> list[np.ndarray]:
+def wiggle_transform(image, intensity=2) -> list[np.ndarray]:
    '''
    返回一组图像振动的帧组，模拟空调运作时的抖动效果
    '''
    frames = []
    height, width = image.shape[:2]
-    shifts = [(-2, 0), (2, 0), (0, -2), (0, 2), (0, 0)]
+    shifts = [(-intensity, 0), (intensity, 0), (0, -intensity), (0, intensity), (0, 0)]
    for dx, dy in shifts:
        M = np.float32([[1, 0, dx], [0, 1, dy]])
        shifted = cv2.warpAffine(image, M, (width, height))
@ -193,7 +254,7 @@ async def generate_ac_image(ac: AirConditioner) -> BytesIO:
        return output
    # 根据生成温度文本图像
-    text = f"{ac.temperature}°C"
+    text = f"{round(ac.temperature, 1)}°C"
    text_image = text_to_transparent_image(
        text, 
        font_size=60, 
@ -218,8 +279,10 @@ async def generate_ac_image(ac: AirConditioner) -> BytesIO:
    final_image_simple = blend_with_transparency(ac_image, transformed_text, (0, 0))
-    frames = wiggle_transform(final_image_simple)
+    intensity = max(2, abs(int(ac.temperature) - 24) // 2)
    frames = wiggle_transform(final_image_simple, intensity=intensity)
    pil_frames = [Image.fromarray(frame) for frame in frames]
    output = BytesIO()
-    pil_frames[0].save(output, format="GIF", save_all=True, append_images=pil_frames[1:], loop=0, duration=50)
+    pil_frames[0].save(output, format="GIF", save_all=True, append_images=pil_frames[1:], loop=0, duration=50, disposal=2)
    return output
--- a/konabot/plugins/idiomgame/init.py
+++ b/konabot/plugins/idiomgame/init.py
@ -30,7 +30,7 @@ def load_banned_ids() -> list[str]:
    if not DATA_FILE_PATH.exists():
        return []
    try:
-        return json.loads(DATA_FILE_PATH.read_text())
+        return json.loads(DATA_FILE_PATH.read_text("utf-8"))
    except Exception as e:
        logger.warning(f"在解析成语接龙封禁文件时遇到问题：{e}")
        return []
@ -45,14 +45,14 @@ def add_banned_id(group_id: str):
    banned_ids = load_banned_ids()
    if group_id not in banned_ids:
        banned_ids.append(group_id)
-        DATA_FILE_PATH.write_text(json.dumps(banned_ids, ensure_ascii=False, indent=4))
+        DATA_FILE_PATH.write_text(json.dumps(banned_ids, ensure_ascii=False, indent=4), "utf-8")
 def remove_banned_id(group_id: str):
    banned_ids = load_banned_ids()
    if group_id in banned_ids:
        banned_ids.remove(group_id)
-        DATA_FILE_PATH.write_text(json.dumps(banned_ids, ensure_ascii=False, indent=4))
+        DATA_FILE_PATH.write_text(json.dumps(banned_ids, ensure_ascii=False, indent=4), "utf-8")
 class TryStartState(Enum):
--- a/konabot/plugins/kona_ph/init.py
+++ b/konabot/plugins/kona_ph/init.py
@ -1,39 +1,57 @@
 import datetime
 import re
 from math import ceil
 from loguru import logger
 from nonebot_plugin_alconna import Alconna, Args, UniMessage, on_alconna
 from konabot.common.nb.qq_broadcast import qq_broadcast
 from konabot.plugins.kona_ph.core.storage import get_today_date
 from konabot.plugins.kona_ph.manager import PUZZLE_PAGE_SIZE, create_admin_commands, config, puzzle_manager
 from konabot.common.longtask import DepLongTaskTarget
 from loguru import logger
 from nonebot import on_message
 from nonebot_plugin_alconna import (Alconna, Args, UniMessage, UniMsg,
                                    on_alconna)
 from nonebot_plugin_apscheduler import scheduler
 from konabot.common.longtask import DepLongTaskTarget
 from konabot.common.nb.qq_broadcast import qq_broadcast
 from konabot.plugins.kona_ph.core.message import (get_daily_report,
                                                  get_daily_report_v2,
                                                  get_puzzle_description,
                                                  get_submission_message)
 from konabot.plugins.kona_ph.core.storage import get_today_date
 from konabot.plugins.kona_ph.manager import (PUZZLE_PAGE_SIZE, config,
                                             create_admin_commands,
                                             puzzle_manager)
 create_admin_commands()
 async def is_play_group(target: DepLongTaskTarget):
-    if target.channel_id in config.plugin_puzzle_playgroup:
+    if target.is_private_chat:
        return True
-    if target.target_id in target.channel_id:
+    if target.channel_id in config.plugin_puzzle_playgroup:
        return True
    return False
-cmd_submit = on_alconna(Alconna(
+cmd_submit = on_message(rule=is_play_group)
-    "re:提交(?:答案|题解|[fF]lag)",
+
    Args["flag", str],
 ), rule=is_play_group)
@cmd_submit.handle()
-async def _(flag: str, target: DepLongTaskTarget):
+async def _(msg: UniMsg, target: DepLongTaskTarget):
-    async with puzzle_manager() as manager:
+    txt = msg.extract_plain_text().strip()
-        result = manager.submit(target.target_id, flag)
+    if match := re.match(r"^提交(?:答案|题解|[fF]lag)\s*(?P<submission>.+?)\s*$", txt):
-        await target.send_message(result.get_unimessage())
+        submission: str = match.group("submission")
        async with puzzle_manager() as manager:
            result = manager.submit(target.target_id, submission)
            if isinstance(result, str):
                await target.send_message(result)
            else:
                await target.send_message(get_submission_message(
                    daily_puzzle_info=result.info,
                    submission=result.submission,
                    puzzle=result.puzzle,
                ))
 cmd_query = on_alconna(Alconna(
-    r"re:(?:(?:所以|话)说?)?今天的题目是什么[啊呀哇呢]?(?:\?？)?"
+    r"re:(?:((?:(?:所以|话)说?)?今天的题目是什么[啊呀哇呢]?(?:\?？)?)|今日谜?题目?)"
 ), rule=is_play_group)
@cmd_query.handle()
@ -42,7 +60,20 @@ async def _(target: DepLongTaskTarget):
        p = manager.get_today_puzzle()
        if p is None:
            return await target.send_message("今天无题，改日再来吧！")
-        await target.send_message(p.get_unimessage())
+        await target.send_message(get_puzzle_description(p))
 cmd_query_submission = on_alconna(Alconna(
    "今日答题情况"
 ), rule=is_play_group)
@cmd_query_submission.handle()
 async def _(target: DepLongTaskTarget):
    gid = None
    if re.match(r"^\d+$", target.channel_id):
        gid = int(target.channel_id)
    async with puzzle_manager() as manager:
        await target.send_message(get_daily_report_v2(manager, gid))
 cmd_history = on_alconna(Alconna(
@ -57,15 +88,14 @@ async def _(target: DepLongTaskTarget, index_id: str = "", page: int = 1):
        today = get_today_date()
        if index_id:
            index_id = index_id.removeprefix("#")
-            if index_id == manager.daily_puzzle_of_date.get(today, ""):
+            if index_id not in manager.daily_puzzle:
-                puzzle = manager.puzzle_data[manager.daily_puzzle[index_id].raw_id]
+                return await target.send_message("没有这道题哦")
-                return await target.send_message(puzzle.get_unimessage())
+            puzzle = manager.puzzle_data[manager.daily_puzzle[index_id].raw_id]
-            if index_id in manager.daily_puzzle:
+            msg = get_puzzle_description(
-                puzzle = manager.puzzle_data[manager.daily_puzzle[index_id].raw_id]
+                puzzle,
-                msg = puzzle.get_unimessage()
+                with_answer=(index_id != manager.daily_puzzle_of_date.get(today, "")),
-                msg = msg.text(f"\n\n------\n\n题解：{puzzle.flag}")
+            )
-                return await target.send_message(msg)
+            return await target.send_message(msg)
            return await target.send_message("没有这道题哦")
        msg = UniMessage.text("====== 历史题目清单 ======\n\n")
        puzzles = [
            (manager.puzzle_data[manager.daily_puzzle[i].raw_id], d)
@ -92,10 +122,14 @@ async def _(target: DepLongTaskTarget, index_id: str = "", page: int = 1):
@scheduler.scheduled_job("cron", hour="8")
 async def _():
    async with puzzle_manager() as manager:
        yesterday = get_today_date() - datetime.timedelta(days=1)
        msg2 = get_daily_report(manager, yesterday)
        if msg2 is not None:
            await qq_broadcast(config.plugin_puzzle_playgroup, msg2)
        puzzle = manager.get_today_puzzle()
        if puzzle is not None:
            logger.info(f"找到了题目 {puzzle.raw_id}，发送")
-            await qq_broadcast(config.plugin_puzzle_playgroup, puzzle.get_unimessage())
+            await qq_broadcast(config.plugin_puzzle_playgroup, get_puzzle_description(puzzle))
        else:
            logger.info("自动任务：没有找到题目，跳过")
--- a/konabot/plugins/kona_ph/core/image.py
+++ b/konabot/plugins/kona_ph/core/image.py
@ -0,0 +1,29 @@
 import nanoid
 from konabot.common.path import ASSETS_PATH
 from konabot.plugins.kona_ph.core.path import KONAPH_IMAGE_BASE
 class PuzzleImageManager:
    def read_puzzle_image(self, img_name: str) -> bytes:
        fp = KONAPH_IMAGE_BASE / img_name
        if fp.exists():
            return fp.read_bytes()
        return (ASSETS_PATH / "img" / "other" / "boom.jpg").read_bytes()
    def upload_puzzle_image(self, data: bytes, suffix: str = ".png") -> str:
        id = nanoid.generate(
            "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz",
            21,
        )
        img_name = f"{id}{suffix}"
        (KONAPH_IMAGE_BASE / img_name).write_bytes(data)
        return img_name
    def remove_puzzle_image(self, img_name: str):
        if img_name:
            (KONAPH_IMAGE_BASE / img_name).unlink(True)
 def get_image_manager() -> PuzzleImageManager:
    return PuzzleImageManager()
--- a/konabot/plugins/kona_ph/core/message.py
+++ b/konabot/plugins/kona_ph/core/message.py
@ -0,0 +1,187 @@
 """
 生成各种各样的 Message 的函数集合
 """
 import datetime
 import functools
 import re
 from typing import Any
 from nonebot_plugin_alconna import UniMessage
 from konabot.common.username import get_username
 from konabot.plugins.kona_ph.core.image import get_image_manager
 from konabot.plugins.kona_ph.core.storage import (DailyPuzzleInfo, Puzzle,
                                                  PuzzleManager,
                                                  PuzzleSubmission)
 def get_puzzle_description(puzzle: Puzzle, with_answer: bool = False) -> UniMessage[Any]:
    """
    获取一个谜题的描述
    """
    img_manager = get_image_manager()
    result = UniMessage.text(f"[KonaPH#{puzzle.index_id}] {puzzle.title}")
    result = result.text(f"\n\n{puzzle.content}")
    if puzzle.img_name:
        result = result.text("\n\n").image(
            raw=img_manager.read_puzzle_image(puzzle.img_name)
        )
    result = result.text(f"\n\n出题者：{get_username(puzzle.author_id)}")
    if with_answer:
        result = result.text(f"\n\n题目答案：{puzzle.flag}")
    else:
        result = result.text("\n\n输入「提交答案 答案」来提交你的解答")
    return result
 def get_submission_message(
    puzzle: Puzzle,
    submission: PuzzleSubmission,
    daily_puzzle_info: DailyPuzzleInfo | None = None,
 ) -> str:
    """
    获得提交答案的反馈信息
    """
    if submission.success:
        rank = -1
        if daily_puzzle_info is not None:
            rank = len(daily_puzzle_info.success_users)
        return f"🎉 恭喜你答对了！你是今天第 {rank} 个解出来的！"
    if submission.hint_id >= 0 and (
        hint := puzzle.hints.get(submission.hint_id)
    ) is not None:
        if hint.is_checkpoint:
            hint_msg = "✨ 恭喜！这是本题的中间答案，加油！"
        else:
            hint_msg = "🤔 答错啦！请检查你的答案。"
        return f"{hint_msg}\n\n提示：{hint.message}"
    return "❌ 答错啦！请检查你的答案。"
 def get_daily_report(
    manager: PuzzleManager,
    date: datetime.date,
 ) -> str | None:
    """
    获得某日的提交的报告信息
    """
    index_id = manager.daily_puzzle_of_date.get(date)
    if index_id is None:
        return None
    info = manager.daily_puzzle[index_id]
    puzzle = manager.puzzle_data[info.raw_id]
    msg = f"[KonaPH#{puzzle.index_id}] 「{puzzle.title}」解答报告\n\n"
    if len(info.success_users) == 0:
        msg += "昨日，无人解出此题 😭😭\n\n"
    else:
        msg += f"昨日，共有 {len(info.success_users)} 人解出此题。\n\n"
        msg += "前五名的解答者：\n\n"
        us = [(u, d) for u, d in info.success_users.items()]
        us = sorted(us, key=lambda t: t[1])
        us = us[:5]
        for u, _ in us:
            m = manager.submissions[puzzle.raw_id][u][-1]
            msg += f"- {get_username(u)} 于 {m.time.strftime('%H:%M')}\n"
        msg += "\n"
    msg += f"出题人：{get_username(puzzle.author_id)}"
    return msg
 def get_daily_report_v2(manager: PuzzleManager, gid: int | None = None):
    p = manager.get_today_puzzle()
    if p is None:
        return "今天无题"
    msg = "==== 今日答题情况 ====\n\n"
    subcount = len(functools.reduce(
        lambda x, y: x + y,
        manager.submissions.get(p.raw_id, {}).values(),
        [],
    ))
    info = manager.daily_puzzle[p.index_id]
    msg += (
        f"总体情况：答对 {len(info.success_users)} / "
        f"参与 {len(info.tried_users)} / "
        f"提交 {subcount}\n"
    )
    success_users = sorted(list(info.success_users.items()), key=lambda v: v[1])
    for u, d in success_users:
        uname = u
        if re.match(r"^\d+$", u):
            uname = get_username(int(u), gid)
        t = d.strftime("%H:%M")
        tries = len(manager.submissions[p.raw_id][u])
        msg += f"\n- {uname} [🎉 {t} 完成 | {tries} 提交]"
    for u in info.tried_users - set(info.success_users.keys()):
        uname = u
        if re.match(r"^\d+$", u):
            uname = get_username(int(u), gid)
        tries = len(manager.submissions[p.raw_id][u])
        checkpoints_touched = len(set((
            s.hint_id for s in manager.submissions[p.raw_id][u]
            if (
                s.hint_id >= 0
                and s.hint_id in p.hints
                and p.hints[s.hint_id].is_checkpoint
            )
        )))
        checkpoint_message = ""
        if checkpoints_touched > 0:
            checkpoint_message = f" | 🚩 {checkpoints_touched} 记录点"
        msg += f"\n- {uname} [💦 {tries} 提交{checkpoint_message}]"
    return msg
 def get_puzzle_info_message(manager: PuzzleManager, puzzle: Puzzle) -> UniMessage[Any]:
    image_manager = get_image_manager()
    status = "✅ 已准备，待发布" if puzzle.ready and not manager.is_puzzle_published(puzzle.raw_id) else \
    (f"🟢 已发布: #{puzzle.index_id}" if manager.is_puzzle_published(puzzle.raw_id) else "⚙️ 未准备")
    status_suffix = ""
    if puzzle.raw_id == manager.puzzle_pinned:
        status_suffix += " | 📌 已被管理员置顶"
    msg = UniMessage.text(
        f"--- 谜题信息 ---\n"
        f"Raw ID: {puzzle.raw_id}\n"
        f"出题者: {get_username(puzzle.author_id)} | {puzzle.author_id}\n"
        f"创建时间: {puzzle.created_at.strftime('%Y-%m-%d %H:%M:%S')}\n"
        f"状态: {status}{status_suffix}\n\n"
        f"标题: {puzzle.title}\n"
        f"Flag: {puzzle.flag}\n\n"
        f"{puzzle.content}"
    )
    if puzzle.img_name:
        msg = msg.image(raw=image_manager.read_puzzle_image(puzzle.img_name))
    msg = msg.text(f"\n---------\n使用 `konaph ready {puzzle.raw_id}` 完成编辑")
    return msg
 def get_puzzle_hint_list(puzzle: Puzzle) -> str:
    msg = f"==== {puzzle.title} 提示与中间答案 ====\n"
    if len(puzzle.hints) == 0:
        msg += "\n你没有添加任何中间答案。"
        return msg
    for hint_id, hint in puzzle.hints.items():
        n = {False: "[提示]", True: "[中间答案]"}[hint.is_checkpoint]
        msg += f"\n{n}[{hint_id}] {hint.pattern}"
        msg += f"\n  {hint.message}"
    return msg
--- a/konabot/plugins/kona_ph/core/path.py
+++ b/konabot/plugins/kona_ph/core/path.py
@ -0,0 +1,9 @@
 from konabot.common.path import DATA_PATH
 KONAPH_BASE = DATA_PATH / "KonaPH"
 KONAPH_DATA_JSON = KONAPH_BASE / "data.json"
 KONAPH_IMAGE_BASE = KONAPH_BASE / "imgs"
 # 保证所有文件夹存在
 KONAPH_BASE.mkdir(exist_ok=True)
 KONAPH_IMAGE_BASE.mkdir(exist_ok=True)
--- a/konabot/plugins/kona_ph/core/storage.py
+++ b/konabot/plugins/kona_ph/core/storage.py
@ -1,26 +1,26 @@
 import asyncio
 import datetime
 import random
-
+import re
 from contextlib import asynccontextmanager
 from pathlib import Path
 from typing import Any
 import nanoid
 from nonebot_plugin_alconna import UniMessage
 from pydantic import BaseModel, Field, ValidationError
-from konabot.common.path import DATA_PATH
+from konabot.plugins.kona_ph.core.path import KONAPH_DATA_JSON
-KONAPH_BASE = DATA_PATH / "KonaPH"
+class PuzzleHint(BaseModel):
-KONAPH_DATA_JSON = KONAPH_BASE / "data.json"
+    pattern: str
-KONAPH_IMAGE_BASE = KONAPH_BASE / "imgs"
+    message: str
    is_checkpoint: bool
-# 保证所有文件夹存在
+
-KONAPH_BASE.mkdir(exist_ok=True)
+class PuzzleSubmission(BaseModel):
-KONAPH_IMAGE_BASE.mkdir(exist_ok=True)
+    success: bool
    flag: str
    time: datetime.datetime
    hint_id: int = -1
 class Puzzle(BaseModel):
@ -37,46 +37,49 @@ class Puzzle(BaseModel):
    flag: str
    ready: bool = False
    published: bool = False
    pinned: bool = False
    created_at: datetime.datetime = Field(default_factory=datetime.datetime.now)
-    def get_image_path(self) -> Path:
+    hints: dict[int, PuzzleHint] = Field(default_factory=dict)
        return KONAPH_IMAGE_BASE / self.img_name
-    def get_unimessage(self) -> UniMessage[Any]:
+    @property
-        result = UniMessage.text(f"[KonaPH#{self.index_id}] {self.title}")
+    def hint_id_max(self) -> int:
-        result = result.text(f"\n\n{self.content}")
+        return max((0, *self.hints.keys()))
-        if self.img_name:
+    def check_submission(
-            result = result.text("\n\n").image(raw=self.get_image_path().read_bytes())
+        self,
-
+        submission: str,
-        result = result.text("\n\n出题者：").at(self.author_id)
+        time: datetime.datetime | None = None,
-        result = result.text("\n\n输入「提交答案 答案」来提交你的解答")
+    ) -> PuzzleSubmission:
-
+        if time is None:
-        return result
+            time = datetime.datetime.now()
-
+        if submission == self.flag:
-    def add_image(self, img: bytes, suffix: str = ".png"):
+            return PuzzleSubmission(
-        if self.img_name:
+                success=True,
-            self.get_image_path().unlink(True)
+                flag=submission,
-        img_id = nanoid.generate(
+                time=time,
-            "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz",
+            )
-            21,
+        for hint_id, hint in self.hints.items():
            if hint.pattern.startswith('/') and hint.pattern.endswith('/'):
                if re.match(hint.pattern.strip('/'), submission):
                    return PuzzleSubmission(
                        success=False,
                        flag=submission,
                        time=time,
                        hint_id=hint_id,
                    )
            else:
                if hint.pattern == submission:
                    return PuzzleSubmission(
                        success=False,
                        flag=submission,
                        time=time,
                        hint_id=hint_id,
                    )
        return PuzzleSubmission(
            success=False,
            flag=submission,
            time=time,
        )
        self.img_name = f"{img_id}{suffix}"
        self.get_image_path().write_bytes(img)
    def remove_image(self):
        if self.img_name:
            self.get_image_path().unlink(True)
            self.img_name = ""
 class PuzzleSubmission(BaseModel):
    success: bool
    flag: str
    time: datetime.datetime
 class DailyPuzzleInfo(BaseModel):
@ -86,15 +89,10 @@ class DailyPuzzleInfo(BaseModel):
    success_users: dict[str, datetime.datetime] = {}
-class PuzzleSubmissionResultMessage(BaseModel):
+class PuzzleSubmissionFeedback(BaseModel):
-    success: bool
+    submission: PuzzleSubmission
-    rank: int = -1
+    puzzle: Puzzle
-    message: str = ""
+    info: DailyPuzzleInfo | None = None
    def get_unimessage(self) -> UniMessage[Any]:
        if self.success:
            return UniMessage.text(f"🎉 恭喜你答对了！你是今天第 {self.rank} 个解出来的！")
        return UniMessage.text(self.message)
 def get_today_date() -> datetime.date:
@ -111,68 +109,61 @@ class PuzzleManager(BaseModel):
    daily_puzzle_of_date: dict[datetime.date, str] = {}
    puzzle_pinned: str = ""
    unpublished_puzzles: set[str] = set()
    unready_puzzles: set[str] = set()
    published_puzzles: set[str] = set()
    index_id_counter: int = 1
    submissions: dict[str, dict[str, list[PuzzleSubmission]]] = {}
    last_pubish_date: datetime.date = Field(
        default_factory=lambda: get_today_date() - datetime.timedelta(days=1)
    )
    last_checked_date: datetime.date = Field(
        default_factory=lambda: get_today_date() - datetime.timedelta(days=1)
    )
    @property
    def last_publish_date(self):
        return max(self.daily_puzzle_of_date.keys())
    @property
    def unpublished_puzzles(self):
        return set((
            p.raw_id for p in self.puzzle_data.values()
            if not self.is_puzzle_published(p.raw_id) and p.ready
        ))
    @property
    def unready_puzzles(self):
        return set((
            p.raw_id for p in self.puzzle_data.values()
            if not self.is_puzzle_published(p.raw_id) and not p.ready
        ))
    @property
    def published_puzzles(self):
        return set((
            p.raw_id for p in self.puzzle_data.values()
            if self.is_puzzle_published(p.raw_id)
        ))
    def is_puzzle_published(self, raw_id: str):
        return raw_id in [i.raw_id for i in self.daily_puzzle.values()]
    def publish_puzzle(self, raw_id: str):
        assert raw_id in self.puzzle_data
-        self.unpublished_puzzles -= set(raw_id)
+        today = get_today_date()
-        self.unready_puzzles -= set(raw_id)
+
        p = self.puzzle_data[raw_id]
        p.index_id = str(self.index_id_counter)
        p.ready = True
        p.published = True
        p.pinned = False
        self.puzzle_pinned = ""
-        self.last_pubish_date = get_today_date()
+        self.last_checked_date = today
        self.last_checked_date = self.last_pubish_date
        self.daily_puzzle[p.index_id] = DailyPuzzleInfo(
            raw_id=raw_id,
-            time=self.last_pubish_date,
+            time=today,
        )
-        self.daily_puzzle_of_date[self.last_pubish_date] = p.index_id
+        self.daily_puzzle_of_date[today] = p.index_id
        self.published_puzzles.add(raw_id)
        self.index_id_counter += 1
    def admin_mark_ready(self, raw_id: str, ready: bool = True):
        if raw_id not in self.puzzle_data:
            return
        if ready:
            self.unready_puzzles -= set(raw_id)
            if raw_id not in self.published_puzzles:
                self.unpublished_puzzles.add(raw_id)
            p = self.puzzle_data[raw_id]
            p.ready = True
            p.published = raw_id in self.published_puzzles
        else:
            self.unready_puzzles.add(raw_id)
            self.unpublished_puzzles -= set(raw_id)
            p = self.puzzle_data[raw_id]
            p.ready = False
            p.published = False
            # if p.raw_id == self.puzzle_pinned:
            #     self.puzzle_pinned = ""
    def admin_pin_puzzle(self, raw_id: str):
        if self.puzzle_pinned:
            p = self.puzzle_data.get(self.puzzle_pinned)
            if p is not None:
                p.pinned = False
        if raw_id in self.puzzle_data:
            p = self.puzzle_data[raw_id]
            p.pinned = True
            self.puzzle_pinned = raw_id
        else:
            self.puzzle_pinned = ""
@ -202,41 +193,23 @@ class PuzzleManager(BaseModel):
            return
        return self.daily_puzzle[p.index_id]
-    def submit(self, user: str, flag: str) -> PuzzleSubmissionResultMessage:
+    def submit(self, user: str, flag: str) -> PuzzleSubmissionFeedback | str:
        p = self.get_today_puzzle()
        d = self.get_today_info()
        now = datetime.datetime.now()
        if p is None or d is None:
-            return PuzzleSubmissionResultMessage(
+            return "今天没有题哦，改天再来吧！"
                success=False,
                message="今天没有题哦，改天再来吧！",
            )
        if user in d.success_users:
-            return PuzzleSubmissionResultMessage(
+            return "你今天已经答对过啦！不用重复提交哦！"
                success=False,
                message="你今天已经答对过啦！不用重复提交哦！",
            )
        if flag != p.flag:
            d.tried_users.add(user)
            self.submissions.setdefault(p.raw_id, {}).setdefault(user, []).append(PuzzleSubmission(
                success=False,
                flag=flag,
                time=now,
            ))
            return PuzzleSubmissionResultMessage(
                success=False,
                message="❌ 答错了，请检查你的答案哦",
            )
        d.tried_users.add(user)
-        d.success_users[user] = now
+        result = p.check_submission(flag, now)
-        self.submissions.setdefault(p.raw_id, {}).setdefault(user, []).append(PuzzleSubmission(
+        self.submissions.setdefault(p.raw_id, {}).setdefault(user, []).append(result)
-            success=True,
+        if result.success:
-            flag=flag,
+            d.success_users[user] = now
-            time=now,
+        return PuzzleSubmissionFeedback(
-        ))
+            submission=result,
-        return PuzzleSubmissionResultMessage(
+            puzzle=p,
-            success=True,
+            info=d,
            rank=len(d.success_users),
        )
    def admin_create_puzzle(self, user: str):
@ -252,9 +225,7 @@ class PuzzleManager(BaseModel):
            author_id=user,
            flag="konaph{this_is_a_flag}",
            ready=False,
            published=False,
        )
        self.unready_puzzles.add(p.raw_id)
        self.puzzle_data[p.raw_id] = p
        return p
@ -264,47 +235,20 @@ class PuzzleManager(BaseModel):
            if p.author_id == user
        ], key=lambda p: p.created_at, reverse=True)
    def get_report_yesterday(self):
        yesterday = get_today_date() - datetime.timedelta(days=1)
        index_id = self.daily_puzzle_of_date.get(yesterday)
        if index_id is None:
            return None
        info = self.daily_puzzle[index_id]
        puzzle = self.puzzle_data[info.raw_id]
        message = UniMessage.text(f"[KonaPH#{index_id}] 「{puzzle.title}」解答报告")
        if len(info.success_users) == 0:
            message = message.text(
                "\n\n昨日，竟无人解出此题！"
            )
        else:
            message = message.text(
                f"\n\n昨日，共有 {len(info.success_users)} 人解出此题。\n\n前五名的解答者："
            )
            us = [(u, d) for u, d in info.success_users.items()]
            us = sorted(us, key=lambda t: t[1])
            us = us[:5]
            for u, _ in us:
                m = self.submissions[puzzle.raw_id][u][-1]
                message = message.text("- ").at(u).text(f" 于 {m.time.strftime('%H:%M')}")
        message = message.text("\n\n出题者：").at(puzzle.author_id)
        return message
 lock = asyncio.Lock()
 def read_data():
    try:
-        data_raw = KONAPH_DATA_JSON.read_text()
+        data_raw = KONAPH_DATA_JSON.read_text("utf-8")
        return PuzzleManager.model_validate_json(data_raw)
    except (FileNotFoundError, ValidationError):
        return PuzzleManager()
 def write_data(data: PuzzleManager):
-    KONAPH_DATA_JSON.write_text(data.model_dump_json())
+    KONAPH_DATA_JSON.write_text(data.model_dump_json(), "utf-8")
@asynccontextmanager
--- a/konabot/plugins/kona_ph/manager.py
+++ b/konabot/plugins/kona_ph/manager.py
@ -1,14 +1,24 @@
 import datetime
 from math import ceil
-from typing import Any
+
 from nonebot import get_plugin_config
-from nonebot_plugin_alconna import Alconna, Args, Image, Option, Query, Subcommand, UniMessage, on_alconna
+from nonebot_plugin_alconna import (Alconna, Args, Image, Option, Query,
                                    Subcommand, SubcommandResult, UniMessage,
                                    on_alconna)
 from pydantic import BaseModel
 from konabot.common.longtask import DepLongTaskTarget
 from konabot.common.nb.exc import BotExceptionMessage
 from konabot.common.nb.extract_image import download_image_bytes
 from konabot.common.nb.qq_broadcast import qq_broadcast
-from konabot.plugins.kona_ph.core.storage import Puzzle, get_today_date, puzzle_manager
+from konabot.common.username import get_username
 from konabot.plugins.kona_ph.core.image import get_image_manager
 from konabot.plugins.kona_ph.core.message import (get_puzzle_description, get_puzzle_hint_list,
                                                  get_puzzle_info_message,
                                                  get_submission_message)
 from konabot.plugins.kona_ph.core.storage import (Puzzle, PuzzleHint, PuzzleManager,
                                                  get_today_date,
                                                  puzzle_manager)
 PUZZLE_PAGE_SIZE = 10
@ -30,31 +40,15 @@ def is_puzzle_admin(target: DepLongTaskTarget):
    return target.target_id in config.plugin_puzzle_admin
-def get_puzzle_info_message(puzzle: Puzzle) -> UniMessage[Any]:
+def check_puzzle(manager: PuzzleManager, target: DepLongTaskTarget, raw_id: str) -> Puzzle:
-    status = "✅ 已准备，待发布" if puzzle.ready and not puzzle.published else \
+    if raw_id not in manager.puzzle_data:
-    (f"🟢 已发布: #{puzzle.index_id}" if puzzle.published else "⚙️ 未准备")
+        raise BotExceptionMessage("没有这个谜题")
-
+    puzzle = manager.puzzle_data[raw_id]
-    status_suffix = ""
+    if is_puzzle_admin(target):
-    if puzzle.pinned:
+        return puzzle
-        status_suffix += " | 📌 已被管理员置顶"
+    if target.target_id != puzzle.author_id:
-
+        raise BotExceptionMessage("你没有权限查看或编辑这个谜题")
-    msg = UniMessage.text(
+    return puzzle
        f"--- 谜题信息 ---\n"
        f"Raw ID: {puzzle.raw_id}\n"
        f"标题: {puzzle.title}\n"
        f"出题者 ID: {puzzle.author_id}\n"
        f"创建时间: {puzzle.created_at.strftime('%Y-%m-%d %H:%M:%S')}\n"
        f"Flag: {puzzle.flag}\n"
        f"状态: {status}{status_suffix}\n\n"
        f"{puzzle.content}"
    )
    if puzzle.img_name:
        msg = msg.image(raw=puzzle.get_image_path().read_bytes())
    msg = msg.text(f"\n---------\n使用 `konaph ready {puzzle.raw_id}` 完成编辑")
    return msg
 def create_admin_commands():
@ -80,6 +74,46 @@ def create_admin_commands():
                dest="modify",
            ),
            Subcommand("publish", Args["raw_id?", str], dest="publish"),
            Subcommand("preview", Args["raw_id", str], dest="preview"),
            Subcommand("get-submits", Args["raw_id", str], dest="get-submits"),
            Subcommand(
                "test",
                Args["raw_id", str],
                Args["submission", str],
                dest="test",
            ),
            Subcommand(
                "hint",
                Subcommand(
                    "add",
                    Args["raw_id", str],
                    Args["pattern", str],
                    Args["message", str],
                    dest="add",
                ),
                Subcommand(
                    "list",
                    Args["raw_id", str],
                    Args["page?", int],
                    dest="list",
                ),
                Subcommand(
                    "modify",
                    Args["raw_id", str],
                    Args["hint_id", int],
                    Option("--pattern", Args["pattern", str], alias=["-p"]),
                    Option("--message", Args["message", str], alias=["-m"]),
                    Option("--checkpoint", Args["is_checkpoint", bool], alias=["-c"]),
                    dest="modify",
                ),
                Subcommand(
                    "delete",
                    Args["raw_id", str],
                    Args["hint_id", int],
                    dest="delete",
                ),
                dest="hint",
            ),
        ),
        rule=is_puzzle_manager,
    )
@ -93,6 +127,10 @@ def create_admin_commands():
        msg = msg.text("konaph info <id> - 查看谜题\n")
        msg = msg.text("konaph my <page?> - 查看我的谜题列表\n")
        msg = msg.text("konaph modify - 查看如何修改谜题信息\n")
        msg = msg.text("konaph preview <id> - 预览一个题目的效果，不会展示答案\n")
        msg = msg.text("konaph get-submits <id> - 获得题目的提交记录\n")
        msg = msg.text("konaph test <id> <answer> - 尝试提交一个答案，看回答的效果\n")
        msg = msg.text("konaph hint - 查看如何编辑题目的中间答案\n")
        if is_puzzle_admin(target):
            msg = msg.text("konaph all [--ready] <page?> - 查看所有谜题\n")
@ -117,20 +155,12 @@ def create_admin_commands():
    @cmd_admin.assign("ready")
    async def _(raw_id: str, target: DepLongTaskTarget):
        async with puzzle_manager() as manager:
-            if raw_id not in manager.puzzle_data:
+            p = check_puzzle(manager, target, raw_id)
                return await target.send_message(UniMessage.text(
                    "你输入的谜题不存在！输入 `konaph my` 查看你创建的谜题"
                ))
            p = manager.puzzle_data[raw_id]
            if p.author_id != target.target_id and not is_puzzle_admin(target):
                return await target.send_message(UniMessage.text(
                    "这不是你的题，你没有权限编辑！输入 `konaph my` 查看你创建的谜题"
                ))
            if p.ready:
                return await target.send_message(UniMessage.text(
                    "题目早就准备好啦！"
                ))
-            manager.admin_mark_ready(raw_id, True)
+            p.ready = True
            await target.send_message(UniMessage.text(
                f"谜题「{p.title}」已经准备就绪！"
            ))
@ -138,25 +168,17 @@ def create_admin_commands():
    @cmd_admin.assign("unready")
    async def _(raw_id: str, target: DepLongTaskTarget):
        async with puzzle_manager() as manager:
-            if raw_id not in manager.puzzle_data:
+            p = check_puzzle(manager, target, raw_id)
                return await target.send_message(UniMessage.text(
                    "你输入的谜题不存在！输入 `konaph my` 查看你创建的谜题"
                ))
            p = manager.puzzle_data[raw_id]
            if p.author_id != target.target_id and not is_puzzle_admin(target):
                return await target.send_message(UniMessage.text(
                    "这不是你的题，你没有权限编辑！输入 `konaph my` 查看你创建的谜题"
                ))
            if not p.ready:
                return await target.send_message(UniMessage.text(
                    f"谜题「{p.title}」已经是未取消状态了！"
                ))
-            if p.published:
+            if manager.is_puzzle_published(p.raw_id):
                return await target.send_message(UniMessage.text(
                    "已发布的谜题不能取消准备状态！"
                ))
-            manager.admin_mark_ready(raw_id, False)
+            p.ready = False
            await target.send_message(UniMessage.text(
                f"谜题「{p.title}」已经取消准备！"
            ))
@ -164,17 +186,8 @@ def create_admin_commands():
    @cmd_admin.assign("info")
    async def _(raw_id: str, target: DepLongTaskTarget):
        async with puzzle_manager() as manager:
-            if raw_id not in manager.puzzle_data:
+            p = check_puzzle(manager, target, raw_id)
-                return await target.send_message(UniMessage.text(
+            await target.send_message(get_puzzle_info_message(manager, p))
                    "你输入的谜题不存在！输入 `konaph my` 查看你创建的谜题"
                ))
            p = manager.puzzle_data[raw_id]
            if p.author_id != target.target_id and not is_puzzle_admin(target):
                return await target.send_message(UniMessage.text(
                    "这不是你的题，你没有权限查看详细信息！"
                ))
            await target.send_message(get_puzzle_info_message(p))
    @cmd_admin.assign("my")
    async def _(target: DepLongTaskTarget, page: int = 1):
@ -193,10 +206,10 @@ def create_admin_commands():
            message = UniMessage.text("==== 我的谜题 ====\n\n")
            for p in puzzles:
                message = message.text("- ")
-                if p.pinned:
+                if manager.puzzle_pinned == p.raw_id:
                    message = message.text("[📌]")
-                if p.published:
+                if manager.is_puzzle_published(p.raw_id):
-                    message = message.text(f"[#{p.index_id}] ")
+                    message = message.text(f"[✨][#{p.index_id}] ")
                elif p.ready:
                    message = message.text("[✅] ")
                else:
@ -207,9 +220,11 @@ def create_admin_commands():
            await target.send_message(message)
    @cmd_admin.assign("all")
-    async def _(target: DepLongTaskTarget, ready: Query[bool] = Query("ready"), page: int = 1):
+    async def _(target: DepLongTaskTarget, ready: Query[bool] = Query("all.ready"), page: int = 1):
        if not is_puzzle_admin(target):
-            return await target.send_message(UniMessage.text("你没有权限查看所有的哦"))
+            return await target.send_message(UniMessage.text(
                "你没有权限使用该指令"
            ))
        async with puzzle_manager() as manager:
            puzzles = [*manager.puzzle_data.values()]
            if ready.available:
@ -224,10 +239,10 @@ def create_admin_commands():
            message = UniMessage.text("==== 所有谜题 ====\n\n")
            for p in puzzles:
                message = message.text("- ")
-                if p.pinned:
+                if p.raw_id == manager.puzzle_pinned:
                    message = message.text("[📌]")
-                if p.published:
+                if manager.is_puzzle_published(p.raw_id):
-                    message = message.text(f"[#{p.index_id}] ")
+                    message = message.text(f"[✨][#{p.index_id}] ")
                elif p.ready:
                    message = message.text("[✅] ")
                else:
@ -276,7 +291,7 @@ def create_admin_commands():
        description: str | None = None,
        flag: str | None = None,
        image: Image | None = None,
-        remove_image: Query[bool] = Query("--remove-image"),
+        remove_image: Query[bool] = Query("modify.remove-image"),
    ):
        if raw_id == "":
            return await target.send_message(
@ -284,35 +299,41 @@ def create_admin_commands():
                "支持的参数：\n"
                "  --title <str> 标题\n"
                "  --description <str> 题目详情描述（用直引号包裹以支持多行）\n"
-                "  --flag <str> flag\n"
+                "  --flag <str> flag，也就是题目的答案\n"
                "  --image <图片> 图片\n"
                "  --remove-image 删除图片"
            )
        image_manager = get_image_manager()
        async with puzzle_manager() as manager:
-            if raw_id not in manager.puzzle_data:
+            p = check_puzzle(manager, target, raw_id)
                return await target.send_message("没有这个谜题")
            p = manager.puzzle_data[raw_id]
            if not is_puzzle_admin(target) and target.target_id != p.author_id:
                return await target.send_message("你没有权限编辑这个谜题")
            if title is not None:
                p.title = title
            if description is not None:
                p.content = description
            if flag is not None:
-                p.flag = flag
+                p.flag = flag.strip()
                if flag.strip() != flag:
                    await target.send_message(
                        "⚠️ 注意：你输入的 Flag 含有开头或结尾的空格，已经帮你去除"
                    )
            if image is not None and image.url is not None:
                b = await download_image_bytes(image.url)
-                p.add_image(b.unwrap())
+                image_manager.remove_puzzle_image(p.img_name)
                p.img_name = image_manager.upload_puzzle_image(b.unwrap())
            elif remove_image.available:
-                p.remove_image()
+                image_manager.remove_puzzle_image(p.img_name)
-            info2 = get_puzzle_info_message(p)
+            info2 = get_puzzle_info_message(manager, p)
            return await target.send_message("修改好啦！看看效果：\n\n" + info2)
    @cmd_admin.assign("publish")
    async def _(target: DepLongTaskTarget, raw_id: str | None = None):
        if not is_puzzle_admin(target):
            return await target.send_message(UniMessage.text(
                "你没有权限使用该指令"
            ))
        today = get_today_date()
        async with puzzle_manager() as manager:
            if today in manager.daily_puzzle_of_date:
@ -323,7 +344,121 @@ def create_admin_commands():
            p = manager.get_today_puzzle(strong=True)
            if p is None:
                return await target.send_message("上架失败了orz，可能是没题了")
-            await qq_broadcast(config.plugin_puzzle_playgroup, p.get_unimessage())
+            await qq_broadcast(config.plugin_puzzle_playgroup, get_puzzle_description(p))
            return await target.send_message("Ok!")
    @cmd_admin.assign("preview")
    async def _(target: DepLongTaskTarget, raw_id: str):
        async with puzzle_manager() as manager:
            p = check_puzzle(manager, target, raw_id)
            return await target.send_message(get_puzzle_description(p))
    @cmd_admin.assign("get-submits")
    async def _(target: DepLongTaskTarget, raw_id: str):
        async with puzzle_manager() as manager:
            puzzle = manager.puzzle_data.get(raw_id)
            if puzzle is None:
                return await target.send_message("没有这个谜题")
            if not is_puzzle_admin(target) and target.target_id != puzzle.author_id:
                return await target.send_message("你没有权限预览这个谜题")
            msg = UniMessage.text(f"==== {puzzle.title} 提交记录 ====\n\n")
            submits = manager.submissions.get(raw_id, {})
            for uid, ls in submits.items():
                s = ', '.join((i.flag for i in ls))
                msg = msg.text(f"- {get_username(uid)}：{s}\n")
            return await target.send_message(msg)
    @cmd_admin.assign("test")
    async def _(target: DepLongTaskTarget, raw_id: str, submission: str):
        """
        测试一道谜题的回答，并给出结果
        """
        async with puzzle_manager() as manager:
            p = check_puzzle(manager, target, raw_id)
            result = p.check_submission(submission)
            msg = get_submission_message(p, result)
            return await target.send_message("[测试提交] " + msg)
    @cmd_admin.assign("subcommands.hint")
    async def _(target: DepLongTaskTarget, subcommands: Query[SubcommandResult] = Query("subcommands.hint")):
        if len(subcommands.result.subcommands) > 0:
            return
        return await target.send_message(
            UniMessage.text("==== 提示/中间答案编辑器 ====\n\n")
            .text("- konaph hint list <id>\n  - 查看某道题的所有提示 / 中间答案\n")
            .text("- konaph hint add <id> <pattern> <hint>\n  - 添加一个提示 / 中间答案\n")
            .text("- konaph hint modify <id> <hint_id>\n")
            .text("  - --pattern <pattern>\n    - 更改匹配规则\n")
            .text("  - --message <message>\n    - 更改提示文本\n")
            .text("  - --checkpoint [True|False]\n    - 更改是否为中间答案\n")
            .text("- konaph hint delete <id> <hint_id>\n  - 删除一个提示 / 中间答案\n")
            .text("\n更多关于 pattern 和中间答案的信息，请见 man：中间答案(7)")
        )
    @cmd_admin.assign("subcommands.hint.add")
    async def _(
        target: DepLongTaskTarget,
        raw_id: str,
        pattern: str,
        message: str,
    ):
        async with puzzle_manager() as manager:
            p = check_puzzle(manager, target, raw_id)
            p.hints[p.hint_id_max + 1] = PuzzleHint(
                pattern=pattern,
                message=message,
                is_checkpoint=False,
            )
            await target.send_message("创建成功！\n\n" + get_puzzle_hint_list(p))
    @cmd_admin.assign("subcommands.hint.list")
    async def _(
        target: DepLongTaskTarget,
        raw_id: str,
    ):
        async with puzzle_manager() as manager:
            p = check_puzzle(manager, target, raw_id)
            await target.send_message(get_puzzle_hint_list(p))
    @cmd_admin.assign("subcommands.hint.modify")
    async def _(
        target: DepLongTaskTarget,
        raw_id: str,
        hint_id: int,
        pattern: str | None = None,
        message: str | None = None,
        is_checkpoint: bool | None = None,
    ):
        async with puzzle_manager() as manager:
            p = check_puzzle(manager, target, raw_id)
            if hint_id not in p.hints:
                raise BotExceptionMessage(
                    f"没有这个 hint_id。请使用 konaph hint list {raw_id} 了解 hint 清单"
                )
            hint = p.hints[hint_id]
            if pattern is not None:
                hint.pattern = pattern
            if message is not None:
                hint.message = message
            if is_checkpoint is not None:
                hint.is_checkpoint = is_checkpoint
            await target.send_message("更改成功！\n\n" + get_puzzle_hint_list(p))
    @cmd_admin.assign("subcommands.hint.delete")
    async def _(
        target: DepLongTaskTarget,
        raw_id: str,
        hint_id: int,
    ):
        async with puzzle_manager() as manager:
            p = check_puzzle(manager, target, raw_id)
            if hint_id not in p.hints:
                raise BotExceptionMessage(
                    f"没有这个 hint_id。请使用 konaph hint list {raw_id} 了解 hint 清单"
                )
            del p.hints[hint_id]
            await target.send_message("删除成功！\n\n" + get_puzzle_hint_list(p))
    return cmd_admin
--- a/konabot/plugins/llm_test.py
+++ b/konabot/plugins/llm_test.py
@ -0,0 +1,40 @@
 """
 肥肠危险注意：本文件仅用于开发环境测试 LLM 模块能否正常工作！
 请不要在生产环境启用它！
 """
 import nonebot
 from nonebot_plugin_alconna import Alconna, Args, on_alconna
 from pydantic import BaseModel
 from konabot.common.llm import get_llm
 from konabot.common.longtask import DepLongTaskTarget
 class LLMTestConfig(BaseModel):
    debug_enable_llm_test: bool = False
 config = nonebot.get_plugin_config(LLMTestConfig)
 if config.debug_enable_llm_test:
    cmd = on_alconna(Alconna(
        "debug-ask-llm",
        Args["prompt", str],
    ))
    @cmd.handle()
    async def _(prompt: str, target: DepLongTaskTarget):
        llm = get_llm()
        msg = await llm.chat(
            [
                {"role": "user", "content": prompt}
            ],
            timeout=None,
            max_tokens=1024,
        )
        content = msg.content or ""
        await target.send_message(content)
--- a/konabot/plugins/man/init.py
+++ b/konabot/plugins/man/init.py
@ -1,4 +1,3 @@
 from curses.ascii import isdigit
 from pathlib import Path
 import nonebot
@ -40,7 +39,10 @@ async def _(
    doc: str | None,
    event: nonebot.adapters.Event,
 ):
-    if doc is not None and section is None and all(isdigit(c) for c in doc):
+    if doc is not None and section is None and all(
        ord('0') <= ord(c) <= ord('9')
        for c in doc
    ):
        section = int(doc)
        doc = None
--- a/konabot/plugins/poll/init.py
+++ b/konabot/plugins/poll/init.py
@ -15,7 +15,7 @@ if not POLL_DATA_FILE.exists():
    POLL_DATA_FILE.write_bytes(POLL_TEMPLATE_FILE.read_bytes())
-poll_list = json.loads(POLL_DATA_FILE.read_text())['poll']
+poll_list = json.loads(POLL_DATA_FILE.read_text("utf-8"))['poll']
 async def createpoll(title,qqid,options):
    polllength = len(poll_list)
@ -53,7 +53,7 @@ def writeback():
    # json.dump({'poll':poll_list},file,ensure_ascii=False,sort_keys=True)
    POLL_DATA_FILE.write_text(json.dumps({
        'poll': poll_list,
-    }, ensure_ascii=False, sort_keys=True))
+    }, ensure_ascii=False, sort_keys=True), "utf-8")
 async def pollvote(polnum,optionnum,qqnum):
    optiond = poll_list[polnum]["polldata"]
--- a/konabot/plugins/roll_dice/roll_dice.py
+++ b/konabot/plugins/roll_dice/roll_dice.py
@ -394,7 +394,8 @@ async def generate_dice_image(number: str) -> BytesIO:
               append_images=images[1:],
               duration=frame_durations,
               format='GIF',
-               loop=1)
+               loop=1,
               disposal=2)
    output.seek(0)
    # pil_final.save(output, format='PNG')
    return output
--- a/konabot/plugins/auto_accept.py
+++ b/konabot/plugins/auto_accept.py
@ -4,10 +4,13 @@ from typing import cast
 from loguru import logger
 from nonebot import get_bot, on_request
 import nonebot
 from nonebot.adapters.onebot.v11.event import FriendRequestEvent
 from nonebot.adapters.onebot.v11.bot import Bot as OnebotBot
 from nonebot_plugin_apscheduler import scheduler
 from konabot.common.nb.is_admin import cfg as adminConfig
 from konabot.common.username import manager
 add_request = on_request()
@ -23,3 +26,15 @@ async def _(req: FriendRequestEvent):
        await req.approve(bot)
        logger.info(f"已经自动同意 {req.user_id} 的好友请求")
@scheduler.scheduled_job("cron", minute="*/5")
 async def _():
    logger.info("尝试更新群成员信息")
    await manager.update()
 driver = nonebot.get_driver()
@driver.on_bot_connect
 async def _():
    logger.info("有 Bot 连接，5 秒后试着更新群成员信息")
    await asyncio.sleep(5)
    await manager.update()
--- a/konabot/plugins/simple_notify/init.py
+++ b/konabot/plugins/simple_notify/init.py
@ -59,14 +59,14 @@ def load_notify_config() -> NotifyConfigFile:
    if not DATA_FILE_PATH.exists():
        return NotifyConfigFile()
    try:
-        return NotifyConfigFile.model_validate_json(DATA_FILE_PATH.read_text())
+        return NotifyConfigFile.model_validate_json(DATA_FILE_PATH.read_text("utf-8"))
    except Exception as e:
        logger.warning(f"在解析 Notify 时遇到问题：{e}")
        return NotifyConfigFile()
 def save_notify_config(config: NotifyConfigFile):
-    DATA_FILE_PATH.write_text(config.model_dump_json(indent=4))
+    DATA_FILE_PATH.write_text(config.model_dump_json(indent=4), "utf-8")
@evt.handle()
--- a/poetry.lock
+++ b/poetry.lock
@ -803,6 +803,23 @@ type = "legacy"
 url = "https://pypi.tuna.tsinghua.edu.cn/simple"
 reference = "mirrors"
 [[package]]
 name = "distro"
 version = "1.9.0"
 description = "Distro - an OS platform information API"
 optional = false
 python-versions = ">=3.6"
 groups = ["main"]
 files = [
    {file = "distro-1.9.0-py3-none-any.whl", hash = "sha256:7bffd925d65168f85027d8da9af6bddab658135b840670a223589bc0c8ef02b2"},
    {file = "distro-1.9.0.tar.gz", hash = "sha256:2fa77c6fd8940f116ee1d6b94a2f90b13b5ea8d019b98bc8bafdcabcdd9bdbed"},
 ]
 [package.source]
 type = "legacy"
 url = "https://pypi.tuna.tsinghua.edu.cn/simple"
 reference = "mirrors"
 [[package]]
 name = "exceptiongroup"
 version = "1.3.0"
@ -1333,6 +1350,123 @@ type = "legacy"
 url = "https://pypi.tuna.tsinghua.edu.cn/simple"
 reference = "mirrors"
 [[package]]
 name = "jiter"
 version = "0.11.1"
 description = "Fast iterable JSON parser."
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
 files = [
    {file = "jiter-0.11.1-cp310-cp310-macosx_10_12_x86_64.whl", hash = "sha256:ed58841a491bbbf3f7c55a6b68fff568439ab73b2cce27ace0e169057b5851df"},
    {file = "jiter-0.11.1-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:499beb9b2d7e51d61095a8de39ebcab1d1778f2a74085f8305a969f6cee9f3e4"},
    {file = "jiter-0.11.1-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:b87b2821795e28cc990939b68ce7a038edea680a24910bd68a79d54ff3f03c02"},
    {file = "jiter-0.11.1-cp310-cp310-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:83f6fa494d8bba14ab100417c80e70d32d737e805cb85be2052d771c76fcd1f8"},
    {file = "jiter-0.11.1-cp310-cp310-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:5fbc6aea1daa2ec6f5ed465f0c5e7b0607175062ceebbea5ca70dd5ddab58083"},
    {file = "jiter-0.11.1-cp310-cp310-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:302288e2edc43174bb2db838e94688d724f9aad26c5fb9a74f7a5fb427452a6a"},
    {file = "jiter-0.11.1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:85db563fe3b367bb568af5d29dea4d4066d923b8e01f3417d25ebecd958de815"},
    {file = "jiter-0.11.1-cp310-cp310-manylinux_2_5_i686.manylinux1_i686.whl", hash = "sha256:f1c1ba2b6b22f775444ef53bc2d5778396d3520abc7b2e1da8eb0c27cb3ffb10"},
    {file = "jiter-0.11.1-cp310-cp310-musllinux_1_1_aarch64.whl", hash = "sha256:523be464b14f8fd0cc78da6964b87b5515a056427a2579f9085ce30197a1b54a"},
    {file = "jiter-0.11.1-cp310-cp310-musllinux_1_1_x86_64.whl", hash = "sha256:25b99b3f04cd2a38fefb22e822e35eb203a2cd37d680dbbc0c0ba966918af336"},
    {file = "jiter-0.11.1-cp310-cp310-win32.whl", hash = "sha256:47a79e90545a596bb9104109777894033347b11180d4751a216afef14072dbe7"},
    {file = "jiter-0.11.1-cp310-cp310-win_amd64.whl", hash = "sha256:cace75621ae9bd66878bf69fbd4dfc1a28ef8661e0c2d0eb72d3d6f1268eddf5"},
    {file = "jiter-0.11.1-cp311-cp311-macosx_10_12_x86_64.whl", hash = "sha256:9b0088ff3c374ce8ce0168523ec8e97122ebb788f950cf7bb8e39c7dc6a876a2"},
    {file = "jiter-0.11.1-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:74433962dd3c3090655e02e461267095d6c84f0741c7827de11022ef8d7ff661"},
    {file = "jiter-0.11.1-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:6d98030e345e6546df2cc2c08309c502466c66c4747b043f1a0d415fada862b8"},
    {file = "jiter-0.11.1-cp311-cp311-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:1d6db0b2e788db46bec2cf729a88b6dd36959af2abd9fa2312dfba5acdd96dcb"},
    {file = "jiter-0.11.1-cp311-cp311-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:55678fbbda261eafe7289165dd2ddd0e922df5f9a1ae46d7c79a5a15242bd7d1"},
    {file = "jiter-0.11.1-cp311-cp311-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:6a6b74fae8e40497653b52ce6ca0f1b13457af769af6fb9c1113efc8b5b4d9be"},
    {file = "jiter-0.11.1-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:0a55a453f8b035eb4f7852a79a065d616b7971a17f5e37a9296b4b38d3b619e4"},
    {file = "jiter-0.11.1-cp311-cp311-manylinux_2_5_i686.manylinux1_i686.whl", hash = "sha256:2638148099022e6bdb3f42904289cd2e403609356fb06eb36ddec2d50958bc29"},
    {file = "jiter-0.11.1-cp311-cp311-musllinux_1_1_aarch64.whl", hash = "sha256:252490567a5d990986f83b95a5f1ca1bf205ebd27b3e9e93bb7c2592380e29b9"},
    {file = "jiter-0.11.1-cp311-cp311-musllinux_1_1_x86_64.whl", hash = "sha256:d431d52b0ca2436eea6195f0f48528202100c7deda354cb7aac0a302167594d5"},
    {file = "jiter-0.11.1-cp311-cp311-win32.whl", hash = "sha256:db6f41e40f8bae20c86cb574b48c4fd9f28ee1c71cb044e9ec12e78ab757ba3a"},
    {file = "jiter-0.11.1-cp311-cp311-win_amd64.whl", hash = "sha256:0cc407b8e6cdff01b06bb80f61225c8b090c3df108ebade5e0c3c10993735b19"},
    {file = "jiter-0.11.1-cp311-cp311-win_arm64.whl", hash = "sha256:fe04ea475392a91896d1936367854d346724a1045a247e5d1c196410473b8869"},
    {file = "jiter-0.11.1-cp312-cp312-macosx_10_12_x86_64.whl", hash = "sha256:c92148eec91052538ce6823dfca9525f5cfc8b622d7f07e9891a280f61b8c96c"},
    {file = "jiter-0.11.1-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:ecd4da91b5415f183a6be8f7158d127bdd9e6a3174138293c0d48d6ea2f2009d"},
    {file = "jiter-0.11.1-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:d7e3ac25c00b9275684d47aa42febaa90a9958e19fd1726c4ecf755fbe5e553b"},
    {file = "jiter-0.11.1-cp312-cp312-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:57d7305c0a841858f866cd459cd9303f73883fb5e097257f3d4a3920722c69d4"},
    {file = "jiter-0.11.1-cp312-cp312-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:e86fa10e117dce22c547f31dd6d2a9a222707d54853d8de4e9a2279d2c97f239"},
    {file = "jiter-0.11.1-cp312-cp312-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:ae5ef1d48aec7e01ee8420155d901bb1d192998fa811a65ebb82c043ee186711"},
    {file = "jiter-0.11.1-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:eb68e7bf65c990531ad8715e57d50195daf7c8e6f1509e617b4e692af1108939"},
    {file = "jiter-0.11.1-cp312-cp312-manylinux_2_5_i686.manylinux1_i686.whl", hash = "sha256:43b30c8154ded5845fa454ef954ee67bfccce629b2dea7d01f795b42bc2bda54"},
    {file = "jiter-0.11.1-cp312-cp312-musllinux_1_1_aarch64.whl", hash = "sha256:586cafbd9dd1f3ce6a22b4a085eaa6be578e47ba9b18e198d4333e598a91db2d"},
    {file = "jiter-0.11.1-cp312-cp312-musllinux_1_1_x86_64.whl", hash = "sha256:677cc2517d437a83bb30019fd4cf7cad74b465914c56ecac3440d597ac135250"},
    {file = "jiter-0.11.1-cp312-cp312-win32.whl", hash = "sha256:fa992af648fcee2b850a3286a35f62bbbaeddbb6dbda19a00d8fbc846a947b6e"},
    {file = "jiter-0.11.1-cp312-cp312-win_amd64.whl", hash = "sha256:88b5cae9fa51efeb3d4bd4e52bfd4c85ccc9cac44282e2a9640893a042ba4d87"},
    {file = "jiter-0.11.1-cp312-cp312-win_arm64.whl", hash = "sha256:9a6cae1ab335551917f882f2c3c1efe7617b71b4c02381e4382a8fc80a02588c"},
    {file = "jiter-0.11.1-cp313-cp313-macosx_10_12_x86_64.whl", hash = "sha256:71b6a920a5550f057d49d0e8bcc60945a8da998019e83f01adf110e226267663"},
    {file = "jiter-0.11.1-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:0b3de72e925388453a5171be83379549300db01284f04d2a6f244d1d8de36f94"},
    {file = "jiter-0.11.1-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:cc19dd65a2bd3d9c044c5b4ebf657ca1e6003a97c0fc10f555aa4f7fb9821c00"},
    {file = "jiter-0.11.1-cp313-cp313-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:d58faaa936743cd1464540562f60b7ce4fd927e695e8bc31b3da5b914baa9abd"},
    {file = "jiter-0.11.1-cp313-cp313-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:902640c3103625317291cb73773413b4d71847cdf9383ba65528745ff89f1d14"},
    {file = "jiter-0.11.1-cp313-cp313-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:30405f726e4c2ed487b176c09f8b877a957f535d60c1bf194abb8dadedb5836f"},
    {file = "jiter-0.11.1-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:3217f61728b0baadd2551844870f65219ac4a1285d5e1a4abddff3d51fdabe96"},
    {file = "jiter-0.11.1-cp313-cp313-manylinux_2_5_i686.manylinux1_i686.whl", hash = "sha256:b1364cc90c03a8196f35f396f84029f12abe925415049204446db86598c8b72c"},
    {file = "jiter-0.11.1-cp313-cp313-musllinux_1_1_aarch64.whl", hash = "sha256:53a54bf8e873820ab186b2dca9f6c3303f00d65ae5e7b7d6bda1b95aa472d646"},
    {file = "jiter-0.11.1-cp313-cp313-musllinux_1_1_x86_64.whl", hash = "sha256:7e29aca023627b0e0c2392d4248f6414d566ff3974fa08ff2ac8dbb96dfee92a"},
    {file = "jiter-0.11.1-cp313-cp313-win32.whl", hash = "sha256:f153e31d8bca11363751e875c0a70b3d25160ecbaee7b51e457f14498fb39d8b"},
    {file = "jiter-0.11.1-cp313-cp313-win_amd64.whl", hash = "sha256:f773f84080b667c69c4ea0403fc67bb08b07e2b7ce1ef335dea5868451e60fed"},
    {file = "jiter-0.11.1-cp313-cp313-win_arm64.whl", hash = "sha256:635ecd45c04e4c340d2187bcb1cea204c7cc9d32c1364d251564bf42e0e39c2d"},
    {file = "jiter-0.11.1-cp313-cp313t-macosx_11_0_arm64.whl", hash = "sha256:d892b184da4d94d94ddb4031296931c74ec8b325513a541ebfd6dfb9ae89904b"},
    {file = "jiter-0.11.1-cp313-cp313t-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:aa22c223a3041dacb2fcd37c70dfd648b44662b4a48e242592f95bda5ab09d58"},
    {file = "jiter-0.11.1-cp313-cp313t-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:330e8e6a11ad4980cd66a0f4a3e0e2e0f646c911ce047014f984841924729789"},
    {file = "jiter-0.11.1-cp313-cp313t-win_amd64.whl", hash = "sha256:09e2e386ebf298547ca3a3704b729471f7ec666c2906c5c26c1a915ea24741ec"},
    {file = "jiter-0.11.1-cp313-cp313t-win_arm64.whl", hash = "sha256:fe4a431c291157e11cee7c34627990ea75e8d153894365a3bc84b7a959d23ca8"},
    {file = "jiter-0.11.1-cp314-cp314-macosx_10_12_x86_64.whl", hash = "sha256:0fa1f70da7a8a9713ff8e5f75ec3f90c0c870be6d526aa95e7c906f6a1c8c676"},
    {file = "jiter-0.11.1-cp314-cp314-macosx_11_0_arm64.whl", hash = "sha256:569ee559e5046a42feb6828c55307cf20fe43308e3ae0d8e9e4f8d8634d99944"},
    {file = "jiter-0.11.1-cp314-cp314-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:f69955fa1d92e81987f092b233f0be49d4c937da107b7f7dcf56306f1d3fcce9"},
    {file = "jiter-0.11.1-cp314-cp314-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:090f4c9d4a825e0fcbd0a2647c9a88a0f366b75654d982d95a9590745ff0c48d"},
    {file = "jiter-0.11.1-cp314-cp314-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:bbf3d8cedf9e9d825233e0dcac28ff15c47b7c5512fdfe2e25fd5bbb6e6b0cee"},
    {file = "jiter-0.11.1-cp314-cp314-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:2aa9b1958f9c30d3d1a558b75f0626733c60eb9b7774a86b34d88060be1e67fe"},
    {file = "jiter-0.11.1-cp314-cp314-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:e42d1ca16590b768c5e7d723055acd2633908baacb3628dd430842e2e035aa90"},
    {file = "jiter-0.11.1-cp314-cp314-manylinux_2_5_i686.manylinux1_i686.whl", hash = "sha256:5db4c2486a023820b701a17aec9c5a6173c5ba4393f26662f032f2de9c848b0f"},
    {file = "jiter-0.11.1-cp314-cp314-musllinux_1_1_aarch64.whl", hash = "sha256:4573b78777ccfac954859a6eff45cbd9d281d80c8af049d0f1a3d9fc323d5c3a"},
    {file = "jiter-0.11.1-cp314-cp314-musllinux_1_1_x86_64.whl", hash = "sha256:7593ac6f40831d7961cb67633c39b9fef6689a211d7919e958f45710504f52d3"},
    {file = "jiter-0.11.1-cp314-cp314-win32.whl", hash = "sha256:87202ec6ff9626ff5f9351507def98fcf0df60e9a146308e8ab221432228f4ea"},
    {file = "jiter-0.11.1-cp314-cp314-win_amd64.whl", hash = "sha256:a5dd268f6531a182c89d0dd9a3f8848e86e92dfff4201b77a18e6b98aa59798c"},
    {file = "jiter-0.11.1-cp314-cp314-win_arm64.whl", hash = "sha256:5d761f863f912a44748a21b5c4979c04252588ded8d1d2760976d2e42cd8d991"},
    {file = "jiter-0.11.1-cp314-cp314t-macosx_11_0_arm64.whl", hash = "sha256:2cc5a3965285ddc33e0cab933e96b640bc9ba5940cea27ebbbf6695e72d6511c"},
    {file = "jiter-0.11.1-cp314-cp314t-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:6b572b3636a784c2768b2342f36a23078c8d3aa6d8a30745398b1bab58a6f1a8"},
    {file = "jiter-0.11.1-cp314-cp314t-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:ad93e3d67a981f96596d65d2298fe8d1aa649deb5374a2fb6a434410ee11915e"},
    {file = "jiter-0.11.1-cp314-cp314t-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:a83097ce379e202dcc3fe3fc71a16d523d1ee9192c8e4e854158f96b3efe3f2f"},
    {file = "jiter-0.11.1-cp314-cp314t-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:7042c51e7fbeca65631eb0c332f90c0c082eab04334e7ccc28a8588e8e2804d9"},
    {file = "jiter-0.11.1-cp314-cp314t-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:0a68d679c0e47649a61df591660507608adc2652442de7ec8276538ac46abe08"},
    {file = "jiter-0.11.1-cp314-cp314t-manylinux_2_5_i686.manylinux1_i686.whl", hash = "sha256:a1b0da75dbf4b6ec0b3c9e604d1ee8beaf15bc046fff7180f7d89e3cdbd3bb51"},
    {file = "jiter-0.11.1-cp314-cp314t-musllinux_1_1_aarch64.whl", hash = "sha256:69dd514bf0fa31c62147d6002e5ca2b3e7ef5894f5ac6f0a19752385f4e89437"},
    {file = "jiter-0.11.1-cp314-cp314t-musllinux_1_1_x86_64.whl", hash = "sha256:bb31ac0b339efa24c0ca606febd8b77ef11c58d09af1b5f2be4c99e907b11111"},
    {file = "jiter-0.11.1-cp314-cp314t-win32.whl", hash = "sha256:b2ce0d6156a1d3ad41da3eec63b17e03e296b78b0e0da660876fccfada86d2f7"},
    {file = "jiter-0.11.1-cp314-cp314t-win_amd64.whl", hash = "sha256:f4db07d127b54c4a2d43b4cf05ff0193e4f73e0dd90c74037e16df0b29f666e1"},
    {file = "jiter-0.11.1-cp314-cp314t-win_arm64.whl", hash = "sha256:28e4fdf2d7ebfc935523e50d1efa3970043cfaa161674fe66f9642409d001dfe"},
    {file = "jiter-0.11.1-cp39-cp39-macosx_10_12_x86_64.whl", hash = "sha256:baa99c8db49467527658bb479857344daf0a14dff909b7f6714579ac439d1253"},
    {file = "jiter-0.11.1-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:860fe55fa3b01ad0edf2adde1098247ff5c303d0121f9ce028c03d4f88c69502"},
    {file = "jiter-0.11.1-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:173dd349d99b6feaf5a25a6fbcaf3489a6f947708d808240587a23df711c67db"},
    {file = "jiter-0.11.1-cp39-cp39-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:14ac1dca837514cc946a6ac2c4995d9695303ecc754af70a3163d057d1a444ab"},
    {file = "jiter-0.11.1-cp39-cp39-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:69af47de5f93a231d5b85f7372d3284a5be8edb4cc758f006ec5a1406965ac5e"},
    {file = "jiter-0.11.1-cp39-cp39-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:685f8b3abd3bbd3e06e4dfe2429ff87fd5d7a782701151af99b1fcbd80e31b2b"},
    {file = "jiter-0.11.1-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:6d04afa2d4e5526e54ae8a58feea953b1844bf6e3526bc589f9de68e86d0ea01"},
    {file = "jiter-0.11.1-cp39-cp39-manylinux_2_5_i686.manylinux1_i686.whl", hash = "sha256:1e92b927259035b50d8e11a8fdfe0ebd014d883e4552d37881643fa289a4bcf1"},
    {file = "jiter-0.11.1-cp39-cp39-musllinux_1_1_aarch64.whl", hash = "sha256:e7bd8be4fad8d4c5558b7801770cd2da6c072919c6f247cc5336edb143f25304"},
    {file = "jiter-0.11.1-cp39-cp39-musllinux_1_1_x86_64.whl", hash = "sha256:121381a77a3c85987f3eba0d30ceaca9116f7463bedeec2fa79b2e7286b89b60"},
    {file = "jiter-0.11.1-cp39-cp39-win32.whl", hash = "sha256:160225407f6dfabdf9be1b44e22f06bc293a78a28ffa4347054698bd712dad06"},
    {file = "jiter-0.11.1-cp39-cp39-win_amd64.whl", hash = "sha256:028e0d59bcdfa1079f8df886cdaefc6f515c27a5288dec956999260c7e4a7cfd"},
    {file = "jiter-0.11.1-graalpy311-graalpy242_311_native-macosx_10_12_x86_64.whl", hash = "sha256:e642b5270e61dd02265866398707f90e365b5db2eb65a4f30c789d826682e1f6"},
    {file = "jiter-0.11.1-graalpy311-graalpy242_311_native-macosx_11_0_arm64.whl", hash = "sha256:464ba6d000585e4e2fd1e891f31f1231f497273414f5019e27c00a4b8f7a24ad"},
    {file = "jiter-0.11.1-graalpy311-graalpy242_311_native-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:055568693ab35e0bf3a171b03bb40b2dcb10352359e0ab9b5ed0da2bf1eb6f6f"},
    {file = "jiter-0.11.1-graalpy311-graalpy242_311_native-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:e0c69ea798d08a915ba4478113efa9e694971e410056392f4526d796f136d3fa"},
    {file = "jiter-0.11.1-graalpy312-graalpy250_312_native-macosx_10_12_x86_64.whl", hash = "sha256:0d4d6993edc83cf75e8c6828a8d6ce40a09ee87e38c7bfba6924f39e1337e21d"},
    {file = "jiter-0.11.1-graalpy312-graalpy250_312_native-macosx_11_0_arm64.whl", hash = "sha256:f78d151c83a87a6cf5461d5ee55bc730dd9ae227377ac6f115b922989b95f838"},
    {file = "jiter-0.11.1-graalpy312-graalpy250_312_native-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:c9022974781155cd5521d5cb10997a03ee5e31e8454c9d999dcdccd253f2353f"},
    {file = "jiter-0.11.1-graalpy312-graalpy250_312_native-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:18c77aaa9117510d5bdc6a946baf21b1f0cfa58ef04d31c8d016f206f2118960"},
    {file = "jiter-0.11.1.tar.gz", hash = "sha256:849dcfc76481c0ea0099391235b7ca97d7279e0fa4c86005457ac7c88e8b76dc"},
 ]
 [package.source]
 type = "legacy"
 url = "https://pypi.tuna.tsinghua.edu.cn/simple"
 reference = "mirrors"
 [[package]]
 name = "linkify-it-py"
 version = "2.0.3"
@ -2200,6 +2334,39 @@ type = "legacy"
 url = "https://pypi.tuna.tsinghua.edu.cn/simple"
 reference = "mirrors"
 [[package]]
 name = "openai"
 version = "2.7.1"
 description = "The official Python library for the openai API"
 optional = false
 python-versions = ">=3.8"
 groups = ["main"]
 files = [
    {file = "openai-2.7.1-py3-none-any.whl", hash = "sha256:2f2530354d94c59c614645a4662b9dab0a5b881c5cd767a8587398feac0c9021"},
    {file = "openai-2.7.1.tar.gz", hash = "sha256:df4d4a3622b2df3475ead8eb0fbb3c27fd1c070fa2e55d778ca4f40e0186c726"},
 ]
 [package.dependencies]
 anyio = ">=3.5.0,<5"
 distro = ">=1.7.0,<2"
 httpx = ">=0.23.0,<1"
 jiter = ">=0.10.0,<1"
 pydantic = ">=1.9.0,<3"
 sniffio = "*"
 tqdm = ">4"
 typing-extensions = ">=4.11,<5"
 [package.extras]
 aiohttp = ["aiohttp", "httpx-aiohttp (>=0.1.9)"]
 datalib = ["numpy (>=1)", "pandas (>=1.2.3)", "pandas-stubs (>=1.1.0.11)"]
 realtime = ["websockets (>=13,<16)"]
 voice-helpers = ["numpy (>=2.0.2)", "sounddevice (>=0.5.1)"]
 [package.source]
 type = "legacy"
 url = "https://pypi.tuna.tsinghua.edu.cn/simple"
 reference = "mirrors"
 [[package]]
 name = "opencc"
 version = "1.1.9"
@ -3387,6 +3554,33 @@ type = "legacy"
 url = "https://pypi.tuna.tsinghua.edu.cn/simple"
 reference = "mirrors"
 [[package]]
 name = "tqdm"
 version = "4.67.1"
 description = "Fast, Extensible Progress Meter"
 optional = false
 python-versions = ">=3.7"
 groups = ["main"]
 files = [
    {file = "tqdm-4.67.1-py3-none-any.whl", hash = "sha256:26445eca388f82e72884e0d580d5464cd801a3ea01e63e5601bdff9ba6a48de2"},
    {file = "tqdm-4.67.1.tar.gz", hash = "sha256:f8aef9c52c08c13a65f30ea34f4e5aac3fd1a34959879d7e59e63027286627f2"},
 ]
 [package.dependencies]
 colorama = {version = "*", markers = "platform_system == \"Windows\""}
 [package.extras]
 dev = ["nbval", "pytest (>=6)", "pytest-asyncio (>=0.24)", "pytest-cov", "pytest-timeout"]
 discord = ["requests"]
 notebook = ["ipywidgets (>=6)"]
 slack = ["slack-sdk"]
 telegram = ["requests"]
 [package.source]
 type = "legacy"
 url = "https://pypi.tuna.tsinghua.edu.cn/simple"
 reference = "mirrors"
 [[package]]
 name = "typing-extensions"
 version = "4.15.0"
@ -3978,4 +4172,4 @@ reference = "mirrors"
 [metadata]
 lock-version = "2.1"
 python-versions = ">=3.12,<4.0"
-content-hash = "ec73430f70658a303c47e6f536ccb0863a475f7f25d5334c8766e6149075648c"
+content-hash = "dcb6567ccb9eb6357179dd8b8eaa5fb69373cef0e17d3a49c7c895d289c0d642"
--- a/pyproject.toml
+++ b/pyproject.toml
@ -27,6 +27,7 @@ dependencies = [
  "nanoid (>=2.0.0,<3.0.0)",
  "opencc (>=1.1.9,<2.0.0)",
  "playwright (>=1.55.0,<2.0.0)",
  "openai (>=2.7.1,<3.0.0)",
 ]
 [build-system]
--- a/scripts/watch_filter.py
+++ b/scripts/watch_filter.py
@ -0,0 +1,13 @@
 from pathlib import Path
 from watchfiles import Change
 base = Path(__file__).parent.parent.absolute()
 def filter(change: Change, path: str) -> bool:
    if "__pycache__" in path:
        return False
    if Path(path).absolute().is_relative_to(base / "data"):
        return False
    return True
Author	SHA1	Message	Date
passthem	7c19c52d9f	添加关于 LLM 配置的文档	2025-11-05 20:36:51 +08:00
passthem	a5f4ae9bdc	添加基础的 LLM 支持	2025-11-05 18:40:13 +08:00
Passthem	9320815d3f	修复无法更改图片的问题 All checks were successful continuous-integration/drone/push Build is passing Details continuous-integration/drone/tag Build is passing Details	2025-11-01 20:59:58 +08:00
Passthem	795300cb83	在每日答题情况添加记录点显示 All checks were successful continuous-integration/drone/push Build is passing Details continuous-integration/drone/tag Build is passing Details	2025-11-01 18:42:47 +08:00
Passthem	0231aa04f4	添加中间答案功能 All checks were successful continuous-integration/drone/push Build is passing Details continuous-integration/drone/tag Build is passing Details	2025-11-01 18:29:10 +08:00
Passthem	01fe33eb9f	部分解耦了 konaph 的一些层	2025-11-01 17:52:05 +08:00
Passthem	adfbac7d90	支持正义 utf-8	2025-11-01 13:48:48 +08:00
Passthem	994c1412da	为 Watchfiles 添加更可配置的过滤器 All checks were successful continuous-integration/drone/push Build is passing Details	2025-11-01 12:40:01 +08:00
passthem	8780dfec6f	在 Tag 成功后也进行 ntfy 通知 All checks were successful continuous-integration/drone/push Build is passing Details continuous-integration/drone/tag Build is passing Details	2025-10-30 16:52:55 +08:00
passthem	490d807e7a	添加一些对题解提交空格的情况判定 All checks were successful continuous-integration/drone/push Build is passing Details	2025-10-30 16:48:09 +08:00
passthem	fa208199ab	我不小心多加了一个 s All checks were successful continuous-integration/drone/push Build is passing Details	2025-10-29 22:01:13 +08:00
passthem	38a17f42a3	添加 Ntfy 构建消息的报告 Some checks failed continuous-integration/drone/push Build is failing Details	2025-10-29 21:59:11 +08:00
passthem	37179fc4d7	添加显示提交记录的指令 All checks were successful continuous-integration/drone/push Build is passing Details continuous-integration/drone/tag Build is passing Details	2025-10-27 15:31:12 +08:00
passthem	56e0aabbf3	优化 UX，添加 preview 指令	2025-10-27 15:20:40 +08:00
MixBadGun	ce2b7fd6f6	空调调温优化与排行榜，浏览器添加本地HTML支持 All checks were successful continuous-integration/drone/push Build is passing Details continuous-integration/drone/tag Build is passing Details	2025-10-27 00:04:27 +08:00
MixBadGun	b28f8f85a2	Merge branch 'master' of https://gitea.service.jazzwhom.top/mttu-developers/konabot	2025-10-26 22:49:04 +08:00
passthem	0acffea86d	添加排行榜，优化 UX All checks were successful continuous-integration/drone/push Build is passing Details continuous-integration/drone/tag Build is passing Details	2025-10-26 12:45:29 +08:00
passthem	3e395f8a35	更少的量，更解耦的数据，更健壮的系统 All checks were successful continuous-integration/drone/push Build is passing Details continuous-integration/drone/tag Build is passing Details	2025-10-26 11:56:03 +08:00
passthem	312e203bbe	忘记把这个答题情况通知加上了 All checks were successful continuous-integration/drone/push Build is passing Details continuous-integration/drone/tag Build is passing Details	2025-10-26 04:07:26 +08:00
passthem	f9deabfce0	修复 Query 逻辑 All checks were successful continuous-integration/drone/push Build is passing Details continuous-integration/drone/tag Build is passing Details	2025-10-26 04:03:51 +08:00
MixBadGun	7a20c3fe2f	空调指数概率损坏与空调、骰子gif图的背景优化	2025-10-26 01:06:26 +08:00