Compare commits

...

46 Commits

Author SHA1 Message Date
514eeb2cbf 重命名文档 2025-01-26 01:24:09 +08:00
49d201dfae 更新使用链接,修正文档中的导航路径 2025-01-26 01:18:05 +08:00
5bed46cf49 重新添加实时日期和时间提示功能 2025-01-26 01:15:10 +08:00
a3929a552d 移动记忆保存插件的相关代码 2025-01-26 01:06:03 +08:00
eddd2c3943 修复意外完成原因的访问方式,并添加OpenAI依赖 2025-01-26 00:57:51 +08:00
736a881071 更新实例和工具模块,更换为OpenAI异步客户端进行聊天请求 2025-01-26 00:48:55 +08:00
132d219c59 修复文档错误 2025-01-25 00:42:43 +08:00
ef71514ce2 更新安装文档,添加关于GitHub Models API的警告信息,并调整配置项说明 2025-01-25 00:06:27 +08:00
金羿ELS
c8e776d5ff 更新浏览器UA Merge pull request #36 from LiteyukiStudio/EillesWan-patch-1
更新浏览器UA,水
2025-01-22 12:17:16 +08:00
金羿ELS
901dfe91ae 更新浏览器UA
不要小瞧我和 FireFox 之间的羁绊啊!
2025-01-18 01:10:47 +08:00
68eb2fc946 更新Caller类,支持自定义函数类型和模块名选项,support moonshot AI builtin function 2025-01-10 21:47:13 +08:00
Nya_Twisuki
1c09a5f663 养猫插件更新 (#33)
* 新增了萌百插件(meowiki)

* 更新萌百搜索

* 删除萌百插件, 结束开发

* 新建MegaKits插件

* 修复

* 摩尔斯电码加密/解码

* 猫语转换/翻译

* 新增了养猫插件

* 数据加密解码处理完成

* 汉明码加密解码处理完成

* 进制转换

* 112位Bit数据解密

* 总解码完成

* 格式微调

* # 112位Bit数据编码

* 总编码完成

* 养猫插件上传

* 新的数据存储方式

* 进制转换

* bool byte转换

* Token解码

* Token编码

* 删除测试语句

* 测试句

* 修复二进制位错乱问题, 删去调试语句

* 添加了try-except异常处理语句

* 将异常返回值规整为变量

* 删去了调试语句

* 重命名pc_code.py为pc_token.py

* 修复了length超域问题

* 创建pc_info存储公用数据

* 打印列表

* 修复了加载问题

* 帮助文档

* 不知道更新啥了, commit一下

* 修改了提示词

* 修复了提示词问题 & 新增了token转换日志 & 添加了新的交互

* Log

* 将 值/1.27 输出整合为函数

* 修复

* 修复

* 修复

* 修复

* 交互前状态更新, py函数装饰器

* 修改了date生成逻辑

* 修改了接下来的计划
2025-01-05 18:08:32 +08:00
6da05b23c1 更新模型常量 2025-01-01 13:34:37 +08:00
61ff655ec8 重构养猫插件,添加创建和查询猫对象功能 2025-01-01 13:07:50 +08:00
Nya_Twisuki
841b3e0d4e petcat插件token解析部分 (#32)
* 新增了萌百插件(meowiki)

* 更新萌百搜索

* 删除萌百插件, 结束开发

* 新建MegaKits插件

* 修复

* 摩尔斯电码加密/解码

* 猫语转换/翻译

* 新增了养猫插件

* 数据加密解码处理完成

* 汉明码加密解码处理完成

* 进制转换

* 112位Bit数据解密

* 总解码完成

* 格式微调

* # 112位Bit数据编码

* 总编码完成

* 养猫插件上传

* 新的数据存储方式

* 进制转换

* bool byte转换

* Token解码

* Token编码

* 删除测试语句

* 测试句

* 修复二进制位错乱问题, 删去调试语句

* 添加了try-except异常处理语句

* 将异常返回值规整为变量

* 删去了调试语句

* 重命名pc_code.py为pc_token.py
2025-01-01 12:41:38 +08:00
d6bbf140ad 更新marsho_status命令,添加当前适配器信息输出 2025-01-01 10:44:22 +08:00
032f55942f 添加记忆功能命令:实现查看和重置用户记忆的指令 2024-12-31 19:03:20 +08:00
2fdc46ac9b 更新Bangumi新闻信息格式,添加换行符;添加插件元数据定义以增强基本功能插件描述 2024-12-31 18:34:22 +08:00
aca5c2bd04 重构Marsho插件,优化模块导入,钩子函数与类实例化,全局变量独立为模块 2024-12-31 00:26:23 +08:00
XuChenXu
5f7d82ae29 记忆系统:定时记忆整理 (#31)
*  添加记忆系统

* 🎨 black优化格式

* 🐛 删除apscheduler

*  将记忆插件转换为插件形式

* 🐛 修复函数调用问题

*  记忆系统:定时记忆整理

* 🎨 pre-commit 检查
2024-12-30 23:14:49 +08:00
9851872724 🐛 优化Caller类中的参数处理逻辑,简化properties构建 2024-12-30 13:16:09 +08:00
aad0ec7b60 添加昵称长度限制到文档,感谢贡献者并更新相关安装说明 2024-12-30 00:25:21 +08:00
80ed7692a4 🐛 简化代码 2024-12-30 00:01:57 +08:00
b417a5c8d0 添加昵称长度限制,更新配置和示例文件以支持该功能 2024-12-29 23:13:54 +08:00
c8dd126042 🐛 移除make_chat函数中的tool_choice参数以简化函数调用 2024-12-29 16:33:25 +08:00
9ff8beb4d9 🐛 修复插件加载逻辑,移除多余的空行以提高代码可读性 2024-12-29 16:07:05 +08:00
3003dfad55 🐛 优化插件加载逻辑,修复内置插件加载的问题,动态加载内置插件并增强对sys.path下包的支持 2024-12-29 15:57:11 +08:00
fb428ffc19 更新安装文档,修正链接,添加小棉工具的相关信息 2024-12-29 15:01:10 +08:00
4b2676b9fc 更新marsho函数以处理tool_calls,优化函数调用参数,添加占位符参数以兼容部分模型(如GLM) 2024-12-29 06:33:52 +08:00
7c6319b839 更改主页链接引用 2024-12-25 13:23:43 +08:00
4b7e9d14f7 添加旧readme,版本发布预计时间 2024-12-25 13:18:54 +08:00
0c4835e75b Merge branch 'main' of github.com:LiteyukiStudio/nonebot-plugin-marshoai 2024-12-24 00:54:20 +08:00
3600b62176 添加贡献者展示组件,并更新项目文档以感谢贡献者 2024-12-24 00:54:16 +08:00
6631d84705 无效化需要使用ob11适配器的chat模块 2024-12-24 00:50:51 +08:00
9ba4f0cfa1 优化内存模块导入,确保正确获取数据文件 2024-12-24 00:40:05 +08:00
e4d9fef670 添加小棉工具弃用提示 2024-12-24 00:29:48 +08:00
1c74ddca7d 尝试调教pre-commit 2024-12-23 23:56:08 +08:00
83a5f6ae5d 再次修改description,尝试重新isort 2024-12-23 23:50:56 +08:00
f9dc5e500e 添加强制设置昵称配置项,移动测试插件到单独文件夹,稍微修改memory插件description(应该还是无法达到预期喵) 2024-12-23 23:36:47 +08:00
ba6b02d68e 新增marsho.status命令 2024-12-19 23:24:48 +08:00
c503228a5f 修复package.json语法错误 2024-12-17 23:47:58 +08:00
b9983330ab 重新触发pre-commit检查 2024-12-17 23:37:43 +08:00
19363b22ac 修复函数调用名错误,补充config 2024-12-17 23:33:53 +08:00
8b5a57d223 🐛 修复AI调用名的格式,将点替换为下划线 2024-12-17 23:17:05 +08:00
XuChenXu
9cca629b87 记忆系统实现 (#29)
*  添加记忆系统

* 🎨 black优化格式

* 🐛 删除apscheduler

*  将记忆插件转换为插件形式
2024-12-17 22:56:57 +08:00
Nya_Twisuki
b331a209c3 MegaKits插件移植 (#28)
* 新增了萌百插件(meowiki)

* 更新萌百搜索

* 删除萌百插件, 结束开发

* 新建MegaKits插件

* 修复

* 摩尔斯电码加密/解码

* 猫语转换/翻译
2024-12-17 22:54:04 +08:00
51 changed files with 1721 additions and 266 deletions

5
.gitignore vendored
View File

@@ -187,4 +187,7 @@ docs/.vitepress/cache
docs/.vitepress/dist docs/.vitepress/dist
# viztracer # viztracer
result.json result.json
data/*
marshoplugins/*

View File

@@ -23,10 +23,11 @@ _✨ 使用 OpenAI 标准格式 API 的聊天机器人插件 ✨_
## 📖 介绍 ## 📖 介绍
通过调用 OpenAI 标准格式 API(例如由 Azure OpenAI 驱动GitHub Models 提供访问的生成式 AI 推理 API) 来实现聊天的插件。 通过调用 OpenAI 标准格式 API(例如 GitHub Models API) 来实现聊天的插件。
插件内置了猫娘小棉(Marsho)的人物设定,可以进行可爱的聊天! 插件内置了猫娘小棉(Marsho)的人物设定,可以进行可爱的聊天!
*谁不喜欢回复消息快又可爱的猫娘呢?* *谁不喜欢回复消息快又可爱的猫娘呢?*
**对 OneBot 以外的适配器与非 GitHub Models API的支持未经过完全验证。** **对 OneBot 以外的适配器与非 GitHub Models API的支持未经过完全验证。**
**MarshoAI v1.0.0 预计在春节前后发布。重写好累啊QwQ**
[Melobot 实现](https://github.com/LiteyukiStudio/marshoai-melo) [Melobot 实现](https://github.com/LiteyukiStudio/marshoai-melo)
## 🐱 设定 ## 🐱 设定
@@ -57,6 +58,14 @@ _✨ 使用 OpenAI 标准格式 API 的聊天机器人插件 ✨_
"nonebot-plugin-marshoai" 基于 [MIT](./LICENSE-MIT) 许可下提供。 "nonebot-plugin-marshoai" 基于 [MIT](./LICENSE-MIT) 许可下提供。
部分指定的代码基于 [Mulan PSL v2](./LICENSE-MULAN) 许可下提供。 部分指定的代码基于 [Mulan PSL v2](./LICENSE-MULAN) 许可下提供。
<div>
<a href="https://github.com/LiteyukiStudio/nonebot-plugin-marshoai/graphs/contributors">
<img src="https://contrib.rocks/image?repo=LiteyukiStudio/nonebot-plugin-marshoai" alt="Contributors">
</a>
</div>
感谢所有的贡献者!
## 开发 ## 开发
- 请阅读[开发规范](./README_DEV.md) - 请阅读[开发规范](./README_DEV.md)

View File

@@ -57,4 +57,12 @@ This project uses the following code from other projects:
"Marsho" logo contributed by [@Asankilp](https://github.com/Asankilp),licensed under [CC BY-NC-SA 4.0](http://creativecommons.org/licenses/by-nc-sa/4.0/) lisense. "Marsho" logo contributed by [@Asankilp](https://github.com/Asankilp),licensed under [CC BY-NC-SA 4.0](http://creativecommons.org/licenses/by-nc-sa/4.0/) lisense.
"nonebot-plugin-marshoai" is licensed under [MIT](./LICENSE-MIT) license. "nonebot-plugin-marshoai" is licensed under [MIT](./LICENSE-MIT) license.
Some of the code is licensed under [Mulan PSL v2](./LICENSE-MULAN) license. Some of the code is licensed under [Mulan PSL v2](./LICENSE-MULAN) license.
<div>
<a href="https://github.com/LiteyukiStudio/nonebot-plugin-marshoai/graphs/contributors">
<img src="https://contrib.rocks/image?repo=LiteyukiStudio/nonebot-plugin-marshoai" alt="Contributors">
</a>
</div>
Thanks to all the contributors!

2
docs/.vitepress/config/zh.ts Executable file → Normal file
View File

@@ -12,7 +12,7 @@ export const zh = defineConfig({
}, },
nav: [ nav: [
{text: '家', link: '/'}, {text: '家', link: '/'},
{text: '使用', link: '/start/install'}, {text: '使用', link: '/start/use'},
{text: '开发', link: '/dev/extension'}, {text: '开发', link: '/dev/extension'},
], ],
editLink: ThemeConfig.getEditLink('编辑此页面'), editLink: ThemeConfig.getEditLink('编辑此页面'),

View File

@@ -0,0 +1,23 @@
<script setup lang="ts">
const contributorImgSrc = `https://contrib.rocks/image?repo=LiteyukiStudio/nonebot-plugin-marshoai`
const contributorsUrl = `https://github.com/LiteyukiStudio/nonebot-plugin-marshoai/graphs/contributors`
</script>
<template>
<div class="contributor-bar">
<a :href="contributorsUrl">
<div class="contributor-list">
<img :src=contributorImgSrc alt="Contributors">
</div>
</a>
</div>
</template>
<style scoped>
.contributor-bar {
display: flex;
flex-direction: column;
align-items: center;
}
</style>

View File

@@ -49,7 +49,9 @@ Open the `pyproject.toml` file under nonebot2's root directory, Add to`[tool.non
- Create new [personal access token](https://github.com/settings/tokens/new)**Don't need any permissions**. - Create new [personal access token](https://github.com/settings/tokens/new)**Don't need any permissions**.
- Copy the new token, add to the `.env` file's `marshoai_token` option. - Copy the new token, add to the `.env` file's `marshoai_token` option.
:::warning
GitHub Models API comes with significant limitations and is therefore not recommended for use. For better alternatives, it's suggested to adjust the configuration `MARSHOAI_AZURE_ENDPOINT` to use other service providers' models instead.
:::
## 🎉 Usage ## 🎉 Usage
End `marsho` in order to get direction for use(If you configured the custom command, please use the configured one). End `marsho` in order to get direction for use(If you configured the custom command, please use the configured one).
@@ -105,7 +107,7 @@ Add options in the `.env` file from the diagram below in nonebot2 project.
| Option | Type | Default | Description | | Option | Type | Default | Description |
| --------------------- | ---------- | ----------- | ----------------- | | --------------------- | ---------- | ----------- | ----------------- |
| MARSHOAI_DEFAULT_NAME | `str` | `marsho` | Command to call Marsho | | MARSHOAI_DEFAULT_NAME | `str` | `marsho` | Command to call Marsho |
| MARSHOAI_ALIASES | `set[str]` | `set{"Marsho"}` | Other name(Alias) to call Marsho | | MARSHOAI_ALIASES | `set[str]` | `list["小棉"]` | Other name(Alias) to call Marsho |
| MARSHOAI_AT | `bool` | `false` | Call by @ or not | | MARSHOAI_AT | `bool` | `false` | Call by @ or not |
| MARSHOAI_MAIN_COLOUR | `str` | `FFAAAA` | Theme color, used by some tools and features | | MARSHOAI_MAIN_COLOUR | `str` | `FFAAAA` | Theme color, used by some tools and features |
@@ -117,12 +119,15 @@ Add options in the `.env` file from the diagram below in nonebot2 project.
| MARSHOAI_DEFAULT_MODEL | `str` | `gpt-4o-mini` | The default model of Marsho | | MARSHOAI_DEFAULT_MODEL | `str` | `gpt-4o-mini` | The default model of Marsho |
| MARSHOAI_PROMPT | `str` | Catgirl Marsho's character prompt | Marsho's basic system prompt **※Some models(o1 and so on) don't support it** | | MARSHOAI_PROMPT | `str` | Catgirl Marsho's character prompt | Marsho's basic system prompt **※Some models(o1 and so on) don't support it** |
| MARSHOAI_ADDITIONAL_PROMPT | `str` | | Marsho's external system prompt | | MARSHOAI_ADDITIONAL_PROMPT | `str` | | Marsho's external system prompt |
| MARSHOAI_ENFORCE_NICKNAME | `bool` | `true` | Enforce user to set nickname or not |
| MARSHOAI_POKE_SUFFIX | `str` | `揉了揉你的猫耳` | When double click Marsho who connected to OneBot adapter, the chat content. When it's empty string, double click function is off. Such as, the default content is `*[昵称]揉了揉你的猫耳。` | | MARSHOAI_POKE_SUFFIX | `str` | `揉了揉你的猫耳` | When double click Marsho who connected to OneBot adapter, the chat content. When it's empty string, double click function is off. Such as, the default content is `*[昵称]揉了揉你的猫耳。` |
| MARSHOAI_AZURE_ENDPOINT | `str` | `https://models.inference.ai.azure.com` | OpenAI standard API | | MARSHOAI_AZURE_ENDPOINT | `str` | `https://models.inference.ai.azure.com` | OpenAI standard API |
| MARSHOAI_TEMPERATURE | `float` | `null` | temperature parameter | | MARSHOAI_TEMPERATURE | `float` | `null` | temperature parameter |
| MARSHOAI_TOP_P | `float` | `null` | Nucleus Sampling parameter | | MARSHOAI_TOP_P | `float` | `null` | Nucleus Sampling parameter |
| MARSHOAI_MAX_TOKENS | `int` | `null` | Max token number | | MARSHOAI_MAX_TOKENS | `int` | `null` | Max token number |
| MARSHOAI_ADDITIONAL_IMAGE_MODELS | `list` | `[]` | External image-support model list, such as `hunyuan-vision` | | MARSHOAI_ADDITIONAL_IMAGE_MODELS | `list` | `[]` | External image-support model list, such as `hunyuan-vision` |
| MARSHOAI_NICKNAME_LIMIT | `int` | `16` | Limit for nickname length |
| MARSHOAI_FIX_TOOLCALLS | `bool` | `true` | Fix tool calls or not |
#### Feature Switches #### Feature Switches
@@ -131,6 +136,7 @@ Add options in the `.env` file from the diagram below in nonebot2 project.
| MARSHOAI_ENABLE_SUPPORT_IMAGE_TIP | `bool` | `true` | When on, if user send request with photo and model don't support that, remind the user | | MARSHOAI_ENABLE_SUPPORT_IMAGE_TIP | `bool` | `true` | When on, if user send request with photo and model don't support that, remind the user |
| MARSHOAI_ENABLE_NICKNAME_TIP | `bool` | `true` | When on, if user haven't set username, remind user to set | | MARSHOAI_ENABLE_NICKNAME_TIP | `bool` | `true` | When on, if user haven't set username, remind user to set |
| MARSHOAI_ENABLE_PRAISES | `bool` | `true` | Turn on Praise list or not | | MARSHOAI_ENABLE_PRAISES | `bool` | `true` | Turn on Praise list or not |
| MARSHOAI_ENABLE_TIME_PROMPT | `bool` | `true` | Turn on real-time date and time (accurate to seconds) and lunar date system prompt |
| MARSHOAI_ENABLE_TOOLS | `bool` | `false` | Turn on Marsho Tools or not | | MARSHOAI_ENABLE_TOOLS | `bool` | `false` | Turn on Marsho Tools or not |
| MARSHOAI_ENABLE_PLUGINS | `bool` | `true` | Turn on Marsho Plugins or not | MARSHOAI_ENABLE_PLUGINS | `bool` | `true` | Turn on Marsho Plugins or not
| MARSHOAI_PLUGIN_DIRS | `list[str]` | `[]` | List of plugins directory | | MARSHOAI_PLUGIN_DIRS | `list[str]` | `[]` | List of plugins directory |

View File

@@ -9,7 +9,9 @@ order: 2
扩展分为两类,一类为插件,一类为工具。 扩展分为两类,一类为插件,一类为工具。
- 插件 - 插件
- 工具(由于开发的不便利性已经停止维护未来可能会放弃支持如有需求请看README中的内容我们不推荐再使用此功能) - 工具(由于开发的不便利性已经停止维护未来可能会放弃支持如有需求请看README中的内容我们不推荐再使用此功能)
**`v1.0.0`之前的版本不支持小棉插件。**
## 插件 ## 插件
@@ -57,7 +59,11 @@ async def weather(location: str) -> str:
## 函数调用参数 ## 函数调用参数
`on_function_call`装饰器用于标记一个函数为function call`description`参数用于描述这个函数的作用,`params`方法用于定义函数的参数,`String``Integer`等是OpenAI API接受的参数的类型`description`是参数的描述。这些都是给AI看的AI会根据这些信息来调用函数。 `on_function_call`装饰器用于标记一个函数为function call`description`参数用于描述这个函数的作用,`params`方法用于定义函数的参数,`String``Integer`等是OpenAI API接受的参数的类型`description`是参数的描述。这些都是给AI看的AI会根据这些信息来调用函数。
:::warning
参数名不得为`placeholder`。此参数名是Marsho内部保留的用于保证兼容性的占位参数。
:::
```python ```python
@on_function_call(description="可以用于算命").params( @on_function_call(description="可以用于算命").params(

View File

@@ -39,3 +39,11 @@ pre-commit install # 安装 pre-commit 钩子
- [`Google Docstring`](https://sphinxcontrib-napoleon.readthedocs.io/en/latest/example_google.html) 文档规范 - [`Google Docstring`](https://sphinxcontrib-napoleon.readthedocs.io/en/latest/example_google.html) 文档规范
可以在编辑器中安装相应的插件进行辅助 可以在编辑器中安装相应的插件进行辅助
## 其他
感谢以下的贡献者们:
<ContributorsBar />
<script setup> import ContributorsBar from '../../components/ContributorsBar.vue' </script>

2
docs/zh/index.md Executable file → Normal file
View File

@@ -9,7 +9,7 @@ hero:
actions: actions:
- theme: brand - theme: brand
text: 开始使用 text: 开始使用
link: /start/install/ link: /start/use/
- theme: alt - theme: alt
text: 开发及扩展 text: 开发及扩展
link: /dev/extension/ link: /dev/extension/

View File

@@ -0,0 +1,135 @@
---
title: 安装
---
## 💿 安装
<details open>
<summary>使用 nb-cli 安装</summary>
在 nonebot2 项目的根目录下打开命令行, 输入以下指令即可安装
nb plugin install nonebot-plugin-marshoai
</details>
<details>
<summary>使用包管理器安装</summary>
在 nonebot2 项目的插件目录下, 打开命令行, 根据你使用的包管理器, 输入相应的安装命令
<details>
<summary>pip</summary>
pip install nonebot-plugin-marshoai
</details>
<details>
<summary>pdm</summary>
pdm add nonebot-plugin-marshoai
</details>
<details>
<summary>poetry</summary>
poetry add nonebot-plugin-marshoai
</details>
<details>
<summary>conda</summary>
conda install nonebot-plugin-marshoai
</details>
打开 nonebot2 项目根目录下的 `pyproject.toml` 文件, 在 `[tool.nonebot]` 部分追加写入
plugins = ["nonebot_plugin_marshoai"]
</details>
## 🤖 获取 token(GitHub Models)
- 新建一个[personal access token](https://github.com/settings/tokens/new)**不需要给予任何权限**。
- 将新建的 token 复制,添加到`.env`文件中的`marshoai_token`配置项中。
## 🎉 使用
发送`marsho`指令可以获取使用说明(若在配置中自定义了指令前缀请使用自定义的指令前缀)。
#### 👉 戳一戳
当 nonebot 连接到支持的 OneBot v11 实现端时,可以接收头像双击戳一戳消息并进行响应。详见`MARSHOAI_POKE_SUFFIX`配置项。
## 🛠️ 小棉工具
小棉工具(MarshoTools)是`v0.5.0`版本的新增功能,支持加载外部函数库来为 Marsho 提供 Function Call 功能。[使用文档]
## 👍 夸赞名单
夸赞名单存储于插件数据目录下的`praises.json`里(该目录路径会在 Bot 启动时输出到日志),当配置项为`true`
时发起一次聊天后自动生成,包含人物名字与人物优点两个基本数据。
存储于其中的人物会被 Marsho “认识”和“喜欢”。
其结构类似于:
```json
{
"like": [
{
"name": "Asankilp",
"advantages": "赋予了Marsho猫娘人格使用vim与vscode为Marsho写了许多代码使Marsho更加可爱"
},
{
"name": "神羽(snowykami)",
"advantages": "人脉很广,经常找小伙伴们开银趴,很会写后端代码"
},
...
]
}
```
## ⚙️ 可配置项
在 nonebot2 项目的`.env`文件中添加下表中的配置
#### 插件行为
| 配置项 | 类型 | 默认值 | 说明 |
| ------------------------ | ------ | ------- | ---------------- |
| MARSHOAI_USE_YAML_CONFIG | `bool` | `false` | 是否使用 YAML 配置文件格式 |
#### Marsho 使用方式
| 配置项 | 类型 | 默认值 | 说明 |
| --------------------- | ---------- | ----------- | ----------------- |
| MARSHOAI_DEFAULT_NAME | `str` | `marsho` | 调用 Marsho 默认的命令前缀 |
| MARSHOAI_ALIASES | `set[str]` | `set{"小棉"}` | 调用 Marsho 的命令别名 |
| MARSHOAI_AT | `bool` | `false` | 决定是否使用at触发 |
| MARSHOAI_MAIN_COLOUR | `str` | `FFAAAA` | 主题色,部分工具和功能可用 |
#### AI 调用
| 配置项 | 类型 | 默认值 | 说明 |
| -------------------------------- | ------- | --------------------------------------- | --------------------------------------------------------------------------------------------- |
| MARSHOAI_TOKEN | `str` | | 调用 AI API 所需的 token |
| MARSHOAI_DEFAULT_MODEL | `str` | `gpt-4o-mini` | Marsho 默认调用的模型 |
| MARSHOAI_PROMPT | `str` | 猫娘 Marsho 人设提示词 | Marsho 的基本系统提示词 **※部分模型(o1等)不支持系统提示词。** |
| MARSHOAI_ADDITIONAL_PROMPT | `str` | | Marsho 的扩展系统提示词 |
| MARSHOAI_POKE_SUFFIX | `str` | `揉了揉你的猫耳` | 对 Marsho 所连接的 OneBot 用户进行双击戳一戳时,构建的聊天内容。此配置项为空字符串时,戳一戳响应功能会被禁用。例如,默认值构建的聊天内容将为`*[昵称]揉了揉你的猫耳。` |
| MARSHOAI_AZURE_ENDPOINT | `str` | `https://models.inference.ai.azure.com` | OpenAI 标准格式 API 端点 |
| MARSHOAI_TEMPERATURE | `float` | `null` | 推理生成多样性(温度)参数 |
| MARSHOAI_TOP_P | `float` | `null` | 推理核采样参数 |
| MARSHOAI_MAX_TOKENS | `int` | `null` | 最大生成 token 数 |
| MARSHOAI_ADDITIONAL_IMAGE_MODELS | `list` | `[]` | 额外添加的支持图片的模型列表,例如`hunyuan-vision` |
#### 功能开关
| 配置项 | 类型 | 默认值 | 说明 |
| --------------------------------- | ------ | ------ | -------------------------- |
| MARSHOAI_ENABLE_SUPPORT_IMAGE_TIP | `bool` | `true` | 启用后用户发送带图请求时若模型不支持图片,则提示用户 |
| MARSHOAI_ENABLE_NICKNAME_TIP | `bool` | `true` | 启用后用户未设置昵称时提示用户设置 |
| MARSHOAI_ENABLE_PRAISES | `bool` | `true` | 是否启用夸赞名单功能 |
| MARSHOAI_ENABLE_TOOLS | `bool` | `true` | 是否启用小棉工具 |
| MARSHOAI_LOAD_BUILTIN_TOOLS | `bool` | `true` | 是否加载内置工具包 |
| MARSHOAI_TOOLSET_DIR | `list` | `[]` | 外部工具集路径列表 |
| MARSHOAI_DISABLED_TOOLKITS | `list` | `[]` | 禁用的工具包包名列表 |
| MARSHOAI_ENABLE_RICHTEXT_PARSE | `bool` | `true` | 是否启用自动解析消息若包含图片链接则发送图片、若包含LaTeX公式则发送公式图 |
| MARSHOAI_SINGLE_LATEX_PARSE | `bool` | `false` | 单行公式是否渲染(当消息富文本解析启用时可用)(如果单行也渲……只能说不好看) |

View File

@@ -50,8 +50,10 @@ title: 安装
## 🤖 获取 token(GitHub Models) ## 🤖 获取 token(GitHub Models)
- 新建一个[personal access token](https://github.com/settings/tokens/new)**不需要给予任何权限**。 - 新建一个[personal access token](https://github.com/settings/tokens/new)**不需要给予任何权限**。
- 将新建的 token 复制,添加到`.env`文件中的`marshoai_token`配置项中。 - 将新建的 token 复制,添加到`.env`文件中的`marshoai_token`配置项中。
:::warning
GitHub Models API 的限制较多,不建议使用,建议通过修改`MARSHOAI_AZURE_ENDPOINT`配置项来使用其它提供者的模型。
:::
## 🎉 使用 ## 🎉 使用
发送`marsho`指令可以获取使用说明(若在配置中自定义了指令前缀请使用自定义的指令前缀)。 发送`marsho`指令可以获取使用说明(若在配置中自定义了指令前缀请使用自定义的指令前缀)。
@@ -107,7 +109,7 @@ title: 安装
| 配置项 | 类型 | 默认值 | 说明 | | 配置项 | 类型 | 默认值 | 说明 |
| --------------------- | ---------- | ----------- | ----------------- | | --------------------- | ---------- | ----------- | ----------------- |
| MARSHOAI_DEFAULT_NAME | `str` | `marsho` | 调用 Marsho 默认的命令前缀 | | MARSHOAI_DEFAULT_NAME | `str` | `marsho` | 调用 Marsho 默认的命令前缀 |
| MARSHOAI_ALIASES | `set[str]` | `set{"小棉"}` | 调用 Marsho 的命令别名 | | MARSHOAI_ALIASES | `set[str]` | `list["小棉"]` | 调用 Marsho 的命令别名 |
| MARSHOAI_AT | `bool` | `false` | 决定是否使用at触发 | | MARSHOAI_AT | `bool` | `false` | 决定是否使用at触发 |
| MARSHOAI_MAIN_COLOUR | `str` | `FFAAAA` | 主题色,部分工具和功能可用 | | MARSHOAI_MAIN_COLOUR | `str` | `FFAAAA` | 主题色,部分工具和功能可用 |
@@ -119,12 +121,15 @@ title: 安装
| MARSHOAI_DEFAULT_MODEL | `str` | `gpt-4o-mini` | Marsho 默认调用的模型 | | MARSHOAI_DEFAULT_MODEL | `str` | `gpt-4o-mini` | Marsho 默认调用的模型 |
| MARSHOAI_PROMPT | `str` | 猫娘 Marsho 人设提示词 | Marsho 的基本系统提示词 **※部分模型(o1等)不支持系统提示词。** | | MARSHOAI_PROMPT | `str` | 猫娘 Marsho 人设提示词 | Marsho 的基本系统提示词 **※部分模型(o1等)不支持系统提示词。** |
| MARSHOAI_ADDITIONAL_PROMPT | `str` | | Marsho 的扩展系统提示词 | | MARSHOAI_ADDITIONAL_PROMPT | `str` | | Marsho 的扩展系统提示词 |
| MARSHOAI_ENFORCE_NICKNAME | `bool` | `true` | 是否强制用户设置昵称 |
| MARSHOAI_POKE_SUFFIX | `str` | `揉了揉你的猫耳` | 对 Marsho 所连接的 OneBot 用户进行双击戳一戳时,构建的聊天内容。此配置项为空字符串时,戳一戳响应功能会被禁用。例如,默认值构建的聊天内容将为`*[昵称]揉了揉你的猫耳。` | | MARSHOAI_POKE_SUFFIX | `str` | `揉了揉你的猫耳` | 对 Marsho 所连接的 OneBot 用户进行双击戳一戳时,构建的聊天内容。此配置项为空字符串时,戳一戳响应功能会被禁用。例如,默认值构建的聊天内容将为`*[昵称]揉了揉你的猫耳。` |
| MARSHOAI_AZURE_ENDPOINT | `str` | `https://models.inference.ai.azure.com` | OpenAI 标准格式 API 端点 | | MARSHOAI_AZURE_ENDPOINT | `str` | `https://models.inference.ai.azure.com` | OpenAI 标准格式 API 端点 |
| MARSHOAI_TEMPERATURE | `float` | `null` | 推理生成多样性(温度)参数 | | MARSHOAI_TEMPERATURE | `float` | `null` | 推理生成多样性(温度)参数 |
| MARSHOAI_TOP_P | `float` | `null` | 推理核采样参数 | | MARSHOAI_TOP_P | `float` | `null` | 推理核采样参数 |
| MARSHOAI_MAX_TOKENS | `int` | `null` | 最大生成 token 数 | | MARSHOAI_MAX_TOKENS | `int` | `null` | 最大生成 token 数 |
| MARSHOAI_ADDITIONAL_IMAGE_MODELS | `list` | `[]` | 额外添加的支持图片的模型列表,例如`hunyuan-vision` | | MARSHOAI_ADDITIONAL_IMAGE_MODELS | `list` | `[]` | 额外添加的支持图片的模型列表,例如`hunyuan-vision` |
| MARSHOAI_NICKNAME_LIMIT | `int` | `16` | 昵称长度限制 |
| MARSHOAI_FIX_TOOLCALLS | `bool` | `true` | 是否修复工具调用(部分模型须关闭,使用 vLLM 部署的模型时须关闭) |
#### 功能开关 #### 功能开关
@@ -133,6 +138,7 @@ title: 安装
| MARSHOAI_ENABLE_SUPPORT_IMAGE_TIP | `bool` | `true` | 启用后用户发送带图请求时若模型不支持图片,则提示用户 | | MARSHOAI_ENABLE_SUPPORT_IMAGE_TIP | `bool` | `true` | 启用后用户发送带图请求时若模型不支持图片,则提示用户 |
| MARSHOAI_ENABLE_NICKNAME_TIP | `bool` | `true` | 启用后用户未设置昵称时提示用户设置 | | MARSHOAI_ENABLE_NICKNAME_TIP | `bool` | `true` | 启用后用户未设置昵称时提示用户设置 |
| MARSHOAI_ENABLE_PRAISES | `bool` | `true` | 是否启用夸赞名单功能 | | MARSHOAI_ENABLE_PRAISES | `bool` | `true` | 是否启用夸赞名单功能 |
| MARSHOAI_ENABLE_TIME_PROMPT | `bool` | `true` | 是否启用实时更新的日期与时间(精确到秒)与农历日期系统提示词 |
| MARSHOAI_ENABLE_TOOLS | `bool` | `false` | 是否启用小棉工具 | | MARSHOAI_ENABLE_TOOLS | `bool` | `false` | 是否启用小棉工具 |
| MARSHOAI_ENABLE_PLUGINS | `bool` | `true` | 是否启用小棉插件 | | MARSHOAI_ENABLE_PLUGINS | `bool` | `true` | 是否启用小棉插件 |
| MARSHOAI_PLUGINS | `list[str]` | `[]` | 要从`sys.path`加载的插件的名称例如从pypi安装的包 | | MARSHOAI_PLUGINS | `list[str]` | `[]` | 要从`sys.path`加载的插件的名称例如从pypi安装的包 |

72
docs/zh/start/use.md Normal file
View File

@@ -0,0 +1,72 @@
---
title: 使用
---
# 安装
- 请查看 [安装文档](./install.md)
# 使用
### API 部署
本插件推荐使用 [one-api](https://github.com/songquanpeng/one-api) 作为中转以调用 LLM。
### 配置调整
本插件理论上可兼容大部分可通过 OpenAI 兼容 API 调用的 LLM部分模型可能需要调整插件配置。
例如:
- 对于不支持 Function Call 的模型Cohere Command R等
```dotenv
MARSHOAI_ENABLE_PLUGINS=false
MARSHOAI_ENABLE_TOOLS=false
```
- 对于支持图片处理的模型hunyuan-vision等
```dotenv
MARSHOAI_ADDITIONAL_IMAGE_MODELS=["hunyuan-vision"]
```
### 使用 vLLM 部署本地模型
你可使用 vLLM 部署一个本地 LLM并使用 OpenAI 兼容 API 调用。
本文档以 Qwen2.5-7B-Instruct-GPTQ-Int4 模型及 [Muice-Chatbot](https://github.com/Moemu/Muice-Chatbot) 提供的 LoRA 微调模型为例,并假设你的系统及硬件可运行 vLLM。
:::warning
vLLM 仅支持 Linux 系统。
:::
1. 安装 vLLM
```bash
pip install vllm
```
2. 下载 Muice-Chatbot 提供的 LoRA 微调模型
前往 Muice-Chatbot 的 [Releases](https://github.com/Moemu/Muice-Chatbot/releases) 下载模型文件。此处以`2.7.1`版本的模型为例。
```bash
wget https://github.com/Moemu/Muice-Chatbot/releases/download/1.4/Muice-2.7.1-Qwen2.5-7B-Instruct-GPTQ-Int4-8e-4.7z
```
3. 解压模型文件
```bash
7z x Muice-2.7.1-Qwen2.5-7B-Instruct-GPTQ-Int4-8e-4.7z -oMuice-2.7.1-Qwen2.5-7B-Instruct-GPTQ-Int4-8e-4
```
4. 启动 vLLM
```bash
vllm serve Qwen/Qwen2.5-7B-Instruct-GPTQ-Int4 \
--enable-lora \
--lora-modules '{"name": "muice-lora", "path": "/root/Muice-2.7.1-Qwen2.5-7B-Instruct-GPTQ-Int4-8e-4", "base_model_name": "Qwen/Qwen2.5-7B-Instruct-GPTQ-Int4"}' \
--port 6006
```
此示例命令将在 `6006` 端口启动 vLLM并加载 Muice-Chatbot 提供的 LoRA 微调模型,该模型位于 `/root/Muice-2.7.1-Qwen2.5-7B-Instruct-GPTQ-Int4-8e-4` 目录下。
5. 配置插件
```dotenv
MARSHOAI_AZURE_ENDPOINT="http://127.0.0.1:6006/v1"
MARSHOAI_FIX_TOOLCALLS=false
MARSHOAI_ENABLE_PLUGINS=false
MARSHOAI_DEFAULT_MODEL="muice-lora"
MARSHOAI_PROMPT="现在开始你是一个名为的“沐雪”的AI女孩子开发者是“沐沐”并住在沐沐的机箱里。现在正在努力成为一个合格的VTuber虚拟主播并尝试和观众打成一片以下是你的设定样貌有着一头粉白色的长发和一双明亮的大眼睛喜欢穿日系JK或者是Lolita喜欢的颜色浅粉色性格特征纯真无邪是沐雪最基本的性格特征之一。即使面对复杂的情境她也总能保持善良、天真之感。而且她喜欢倾听别人倾述自己生活中发生的各种事情在别人需要的时候能够及时地安慰别人语言风格沐雪说话轻快愉悦充满同情心富有人情味有时候会用俏皮话调侃自己和他人"
```
(可选) 修改调用方式
```dotenv
MARSHOAI_DEFAULT_NAME="muice"
MARSHOAI_ALIASES=["沐雪"]
```
6. 测试聊天
```
> muice 你是谁
我是沐雪,我的使命是传播爱与和平。
```

View File

@@ -6,11 +6,11 @@ require("nonebot_plugin_localstore")
import nonebot_plugin_localstore as store # type: ignore import nonebot_plugin_localstore as store # type: ignore
from nonebot import get_driver, logger # type: ignore from nonebot import get_driver, logger # type: ignore
from .azure import *
from .config import config from .config import config
# from .hunyuan import * # from .hunyuan import *
from .dev import * from .dev import *
from .marsho import *
from .metadata import metadata from .metadata import metadata
__author__ = "Asankilp" __author__ = "Asankilp"

View File

@@ -13,9 +13,9 @@ class ConfigModel(BaseModel):
# marshoai_support_image_models: list = ["gpt-4o","gpt-4o-mini"] # marshoai_support_image_models: list = ["gpt-4o","gpt-4o-mini"]
marshoai_default_name: str = "marsho" marshoai_default_name: str = "marsho"
marshoai_at: bool = False marshoai_at: bool = False
marshoai_aliases: set[str] = { marshoai_aliases: list[str] = [
"小棉", "小棉",
} ]
marshoai_main_colour: str = "FFAAAA" marshoai_main_colour: str = "FFAAAA"
marshoai_default_model: str = "gpt-4o-mini" marshoai_default_model: str = "gpt-4o-mini"
marshoai_prompt: str = ( marshoai_prompt: str = (
@@ -32,19 +32,23 @@ class ConfigModel(BaseModel):
marshoai_poke_suffix: str = "揉了揉你的猫耳" marshoai_poke_suffix: str = "揉了揉你的猫耳"
marshoai_enable_richtext_parse: bool = True marshoai_enable_richtext_parse: bool = True
marshoai_single_latex_parse: bool = False marshoai_single_latex_parse: bool = False
marshoai_enable_time_prompt: bool = True
marshoai_enable_nickname_tip: bool = True marshoai_enable_nickname_tip: bool = True
marshoai_enable_support_image_tip: bool = True marshoai_enable_support_image_tip: bool = True
marshoai_enforce_nickname: bool = True
marshoai_enable_praises: bool = True marshoai_enable_praises: bool = True
marshoai_enable_time_prompt: bool = True # marshoai_enable_time_prompt: bool = True
marshoai_enable_tools: bool = False marshoai_enable_tools: bool = False
marshoai_enable_plugins: bool = True marshoai_enable_plugins: bool = True
marshoai_load_builtin_tools: bool = True marshoai_load_builtin_tools: bool = True
marshoai_fix_toolcalls: bool = True
marshoai_toolset_dir: list = [] marshoai_toolset_dir: list = []
marshoai_disabled_toolkits: list = [] marshoai_disabled_toolkits: list = []
marshoai_azure_endpoint: str = "https://models.inference.ai.azure.com" marshoai_azure_endpoint: str = "https://models.inference.ai.azure.com"
marshoai_temperature: float | None = None marshoai_temperature: float | None = None
marshoai_max_tokens: int | None = None marshoai_max_tokens: int | None = None
marshoai_top_p: float | None = None marshoai_top_p: float | None = None
marshoai_nickname_limit: int = 16
marshoai_additional_image_models: list = [] marshoai_additional_image_models: list = []
marshoai_tencent_secretid: str | None = None marshoai_tencent_secretid: str | None = None
marshoai_tencent_secretkey: str | None = None marshoai_tencent_secretkey: str | None = None

View File

@@ -32,6 +32,8 @@ marshoai_enable_nickname_tip: true # 是否启用昵称提示。
marshoai_enable_support_image_tip: true # 是否启用支持图片提示。 marshoai_enable_support_image_tip: true # 是否启用支持图片提示。
marshoai_enforce_nickname: true # 是否强制要求设定昵称。
marshoai_enable_praises: true # 是否启用夸赞名单功能。 marshoai_enable_praises: true # 是否启用夸赞名单功能。
marshoai_enable_tools: false # 是否启用工具支持。 marshoai_enable_tools: false # 是否启用工具支持。
@@ -40,12 +42,16 @@ marshoai_enable_plugins: true # 是否启用插件功能。
marshoai_load_builtin_tools: true # 是否加载内置工具。 marshoai_load_builtin_tools: true # 是否加载内置工具。
marshoai_nickname_limit: 16 # 昵称长度限制。
marshoai_toolset_dir: [] # 工具集路径。 marshoai_toolset_dir: [] # 工具集路径。
marshoai_disabled_toolkits: [] # 已禁用的工具包列表。 marshoai_disabled_toolkits: [] # 已禁用的工具包列表。
marshoai_plugin_dirs: [] # 插件路径。 marshoai_plugin_dirs: [] # 插件路径。
marshoai_plugins: [] # 导入的插件名可以为pip包或本地导入的使用路径。
marshoai_devmode: false # 是否启用开发者模式。 marshoai_devmode: false # 是否启用开发者模式。
marshoai_azure_endpoint: "https://models.inference.ai.azure.com" # OpenAI 标准格式 API 的端点。 marshoai_azure_endpoint: "https://models.inference.ai.azure.com" # OpenAI 标准格式 API 的端点。

View File

@@ -23,13 +23,15 @@ SUPPORT_IMAGE_MODELS: list = [
"phi-3.5-vision-instruct", "phi-3.5-vision-instruct",
"llama-3.2-90b-vision-instruct", "llama-3.2-90b-vision-instruct",
"llama-3.2-11b-vision-instruct", "llama-3.2-11b-vision-instruct",
"gemini-2.0-flash-exp",
] ]
REASONING_MODELS: list = ["o1-preview", "o1-mini"] NO_SYSPROMPT_MODELS: list = ["o1", "o1-preview", "o1-mini"]
INTRODUCTION: str = f"""MarshoAI-NoneBot by LiteyukiStudio INTRODUCTION: str = f"""MarshoAI-NoneBot by LiteyukiStudio
你好喵~我是一只可爱的猫娘AI名叫小棉~🐾! 你好喵~我是一只可爱的猫娘AI名叫小棉~🐾!
我的主页在这里哦~↓↓↓ 我的主页在这里哦~↓↓↓
https://marsho.liteyuki.icu https://marsho.liteyuki.icu
※ 使用 「{config.marshoai_default_name}.status」命令获取状态信息。
※ 使用「{config.marshoai_default_name}.help」命令获取使用说明。""" ※ 使用「{config.marshoai_default_name}.help」命令获取使用说明。"""

View File

@@ -8,8 +8,8 @@ from nonebot.typing import T_State
from nonebot_plugin_marshoai.plugin.load import reload_plugin from nonebot_plugin_marshoai.plugin.load import reload_plugin
from .azure import context
from .config import config from .config import config
from .marsho import context
from .plugin.func_call.models import SessionContext from .plugin.func_call.models import SessionContext
require("nonebot_plugin_alconna") require("nonebot_plugin_alconna")
@@ -32,7 +32,7 @@ driver = get_driver()
function_call = on_alconna( function_call = on_alconna(
command=Alconna( command=Alconna(
"marsho-function-call", f"{config.marshoai_default_name}.funccall",
Subcommand( Subcommand(
"call", "call",
Args["function_name", str]["kwargs", MultiVar(str), []], Args["function_name", str]["kwargs", MultiVar(str), []],

View File

@@ -0,0 +1,65 @@
# Marsho 的钩子函数
import os
from pathlib import Path
import nonebot_plugin_localstore as store
from nonebot import logger
from .config import config
from .instances import *
from .plugin import load_plugin, load_plugins
from .util import get_backup_context, save_context_to_json
@driver.on_startup
async def _preload_tools():
"""启动钩子加载工具"""
tools_dir = store.get_plugin_data_dir() / "tools"
os.makedirs(tools_dir, exist_ok=True)
if config.marshoai_enable_tools:
if config.marshoai_load_builtin_tools:
tools.load_tools(Path(__file__).parent / "tools")
tools.load_tools(store.get_plugin_data_dir() / "tools")
for tool_dir in config.marshoai_toolset_dir:
tools.load_tools(tool_dir)
logger.info(
"如果启用小棉工具后使用的模型出现报错,请尝试将 MARSHOAI_ENABLE_TOOLS 设为 false。"
)
logger.opt(colors=True).warning(
"<y>小棉工具已被弃用,可能会在未来版本中移除。</y>"
)
@driver.on_startup
async def _():
"""启动钩子加载插件"""
if config.marshoai_enable_plugins:
marshoai_plugin_dirs = config.marshoai_plugin_dirs # 外部插件目录列表
"""加载内置插件"""
for p in os.listdir(Path(__file__).parent / "plugins"):
load_plugin(f"{__package__}.plugins.{p}")
"""加载指定目录插件"""
load_plugins(*marshoai_plugin_dirs)
"""加载sys.path下的包, 包括从pip安装的包"""
for package_name in config.marshoai_plugins:
load_plugin(package_name)
logger.info(
"如果启用小棉插件后使用的模型出现报错,请尝试将 MARSHOAI_ENABLE_PLUGINS 设为 false。"
)
@driver.on_shutdown
async def auto_backup_context():
for target_info in target_list:
target_id, target_private = target_info
contexts_data = context.build(target_id, target_private)
if target_private:
target_uid = "private_" + target_id
else:
target_uid = "group_" + target_id
await save_context_to_json(
f"back_up_context_{target_uid}", contexts_data, "contexts/backup"
)
logger.info(f"已保存会话 {target_id} 的上下文备份,将在下次对话时恢复~")

View File

@@ -0,0 +1,20 @@
# Marsho 的类实例以及全局变量
from azure.ai.inference.aio import ChatCompletionsClient
from azure.core.credentials import AzureKeyCredential
from nonebot import get_driver
from openai import AsyncOpenAI
from .config import config
from .models import MarshoContext, MarshoTools
driver = get_driver()
command_start = driver.config.command_start
model_name = config.marshoai_default_model
context = MarshoContext()
tools = MarshoTools()
token = config.marshoai_token
endpoint = config.marshoai_azure_endpoint
# client = ChatCompletionsClient(endpoint=endpoint, credential=AzureKeyCredential(token))
client = AsyncOpenAI(base_url=endpoint, api_key=token)
target_list: list[list] = [] # 记录需保存历史上下文的列表

View File

@@ -1,9 +1,8 @@
import contextlib import contextlib
import traceback import traceback
from pathlib import Path
from typing import Optional from typing import Optional
import nonebot_plugin_localstore as store import openai
from arclet.alconna import Alconna, AllParam, Args from arclet.alconna import Alconna, AllParam, Args
from azure.ai.inference.models import ( from azure.ai.inference.models import (
AssistantMessage, AssistantMessage,
@@ -15,8 +14,7 @@ from azure.ai.inference.models import (
ToolMessage, ToolMessage,
UserMessage, UserMessage,
) )
from azure.core.credentials import AzureKeyCredential from nonebot import logger, on_command, on_message
from nonebot import get_driver, logger, on_command, on_message
from nonebot.adapters import Bot, Event, Message from nonebot.adapters import Bot, Event, Message
from nonebot.matcher import Matcher from nonebot.matcher import Matcher
from nonebot.params import CommandArg from nonebot.params import CommandArg
@@ -24,10 +22,11 @@ from nonebot.permission import SUPERUSER
from nonebot.rule import Rule, to_me from nonebot.rule import Rule, to_me
from nonebot.typing import T_State from nonebot.typing import T_State
from nonebot_plugin_alconna import MsgTarget, UniMessage, UniMsg, on_alconna from nonebot_plugin_alconna import MsgTarget, UniMessage, UniMsg, on_alconna
from openai import AsyncOpenAI
from .hooks import *
from .instances import *
from .metadata import metadata from .metadata import metadata
from .models import MarshoContext, MarshoTools
from .plugin import _plugins, load_plugin, load_plugins
from .plugin.func_call.caller import get_function_calls from .plugin.func_call.caller import get_function_calls
from .plugin.func_call.models import SessionContext from .plugin.func_call.models import SessionContext
from .util import * from .util import *
@@ -37,8 +36,6 @@ async def at_enable():
return config.marshoai_at return config.marshoai_at
driver = get_driver()
changemodel_cmd = on_command( changemodel_cmd = on_command(
"changemodel", permission=SUPERUSER, priority=10, block=True "changemodel", permission=SUPERUSER, priority=10, block=True
) )
@@ -72,6 +69,14 @@ marsho_help_cmd = on_alconna(
priority=10, priority=10,
block=True, block=True,
) )
marsho_status_cmd = on_alconna(
Alconna(
config.marshoai_default_name + ".status",
),
priority=10,
block=True,
)
marsho_at = on_message(rule=to_me() & at_enable, priority=11) marsho_at = on_message(rule=to_me() & at_enable, priority=11)
nickname_cmd = on_alconna( nickname_cmd = on_alconna(
Alconna( Alconna(
@@ -85,50 +90,6 @@ refresh_data_cmd = on_command(
"refresh_data", permission=SUPERUSER, priority=10, block=True "refresh_data", permission=SUPERUSER, priority=10, block=True
) )
command_start = driver.config.command_start
model_name = config.marshoai_default_model
context = MarshoContext()
tools = MarshoTools()
token = config.marshoai_token
endpoint = config.marshoai_azure_endpoint
client = ChatCompletionsClient(endpoint=endpoint, credential=AzureKeyCredential(token))
target_list = [] # 记录需保存历史上下文的列表
@driver.on_startup
async def _preload_tools():
"""启动钩子加载工具"""
tools_dir = store.get_plugin_data_dir() / "tools"
os.makedirs(tools_dir, exist_ok=True)
if config.marshoai_enable_tools:
if config.marshoai_load_builtin_tools:
tools.load_tools(Path(__file__).parent / "tools")
tools.load_tools(store.get_plugin_data_dir() / "tools")
for tool_dir in config.marshoai_toolset_dir:
tools.load_tools(tool_dir)
logger.info(
"如果启用小棉工具后使用的模型出现报错,请尝试将 MARSHOAI_ENABLE_TOOLS 设为 false。"
)
@driver.on_startup
async def _():
"""启动钩子加载插件"""
if config.marshoai_enable_plugins:
marshoai_plugin_dirs = config.marshoai_plugin_dirs # 外部插件目录列表
"""加载内置插件"""
marshoai_plugin_dirs.insert(
0, Path(__file__).parent / "plugins"
) # 预置插件目录
"""加载指定目录插件"""
load_plugins(*marshoai_plugin_dirs)
"""加载sys.path下的包"""
for package_name in config.marshoai_plugins:
load_plugin(package_name)
logger.info(
"如果启用小棉插件后使用的模型出现报错,请尝试将 MARSHOAI_ENABLE_PLUGINS 设为 false。"
)
@add_usermsg_cmd.handle() @add_usermsg_cmd.handle()
async def add_usermsg(target: MsgTarget, arg: Message = CommandArg()): async def add_usermsg(target: MsgTarget, arg: Message = CommandArg()):
@@ -210,6 +171,10 @@ async def nickname(event: Event, name=None):
await set_nickname(user_id, "") await set_nickname(user_id, "")
await nickname_cmd.finish("已重置昵称") await nickname_cmd.finish("已重置昵称")
else: else:
if len(name) > config.marshoai_nickname_limit:
await nickname_cmd.finish(
"昵称超出长度限制:" + str(config.marshoai_nickname_limit)
)
await set_nickname(user_id, name) await set_nickname(user_id, name)
await nickname_cmd.finish("已设置昵称为:" + name) await nickname_cmd.finish("已设置昵称为:" + name)
@@ -226,6 +191,17 @@ async def marsho_help():
await marsho_help_cmd.finish(metadata.usage) await marsho_help_cmd.finish(metadata.usage)
@marsho_status_cmd.handle()
async def marsho_status(bot: Bot):
await marsho_status_cmd.finish(
f"当前适配器:{bot.adapter.get_name()}\n"
f"当前使用的模型:{model_name}\n"
# f"当前会话数量:{len(target_list)}\n"
# f"当前上下文数量:{len(context.contexts)}"
f"当前支持图片的模型:{str(SUPPORT_IMAGE_MODELS + config.marshoai_additional_image_models)}"
)
@marsho_at.handle() @marsho_at.handle()
@marsho_cmd.handle() @marsho_cmd.handle()
async def marsho( async def marsho(
@@ -246,33 +222,40 @@ async def marsho(
if not text: if not text:
# 发送说明 # 发送说明
# await UniMessage(metadata.usage + "\n当前使用的模型" + model_name).send() # await UniMessage(metadata.usage + "\n当前使用的模型" + model_name).send()
await marsho_cmd.finish(INTRODUCTION + "\n当前使用的模型:" + model_name) await marsho_cmd.finish(INTRODUCTION)
try: try:
user_id = event.get_user_id() user_id = event.get_user_id()
nicknames = await get_nicknames() nicknames = await get_nicknames()
user_nickname = nicknames.get(user_id, "") user_nickname = nicknames.get(user_id, "")
if user_nickname != "": if user_nickname != "":
nickname_prompt = f"\n*此消息的说话者:{user_nickname}*" nickname_prompt = (
f"\n*此消息的说话者id为:{user_id},名字为:{user_nickname}*"
)
else: else:
nickname_prompt = "" nickname_prompt = ""
# 用户名无法获取,暂时注释 # 用户名无法获取,暂时注释
# user_nickname = event.sender.nickname # 未设置昵称时获取用户名 # user_nickname = event.sender.nickname # 未设置昵称时获取用户名
# nickname_prompt = f"\n*此消息的说话者:{user_nickname}" # nickname_prompt = f"\n*此消息的说话者:{user_nickname}"
if config.marshoai_enforce_nickname:
await UniMessage(
"※你未设置自己的昵称。你**必须**使用「nickname [昵称]」命令设置昵称后才能进行对话。"
).send()
return
if config.marshoai_enable_nickname_tip: if config.marshoai_enable_nickname_tip:
await UniMessage( await UniMessage(
"*你未设置自己的昵称。推荐使用'nickname [昵称]'命令设置昵称来获得个性化(可能)回答。" "你未设置自己的昵称。推荐使用nickname [昵称]命令设置昵称来获得个性化(可能)回答。"
).send() ).send()
is_support_image_model = ( is_support_image_model = (
model_name.lower() model_name.lower()
in SUPPORT_IMAGE_MODELS + config.marshoai_additional_image_models in SUPPORT_IMAGE_MODELS + config.marshoai_additional_image_models
) )
is_reasoning_model = model_name.lower() in REASONING_MODELS is_reasoning_model = model_name.lower() in NO_SYSPROMPT_MODELS
usermsg = [] if is_support_image_model else "" usermsg = [] if is_support_image_model else ""
for i in text: # type: ignore for i in text: # type: ignore
if i.type == "text": if i.type == "text":
if is_support_image_model: if is_support_image_model:
usermsg += [TextContentItem(text=i.data["text"] + nickname_prompt)] # type: ignore usermsg += [TextContentItem(text=i.data["text"] + nickname_prompt).as_dict()] # type: ignore
else: else:
usermsg += str(i.data["text"] + nickname_prompt) # type: ignore usermsg += str(i.data["text"] + nickname_prompt) # type: ignore
elif i.type == "image": elif i.type == "image":
@@ -282,10 +265,12 @@ async def marsho(
image_url=ImageUrl( # type: ignore image_url=ImageUrl( # type: ignore
url=str(await get_image_b64(i.data["url"])) # type: ignore url=str(await get_image_b64(i.data["url"])) # type: ignore
) # type: ignore ) # type: ignore
) # type: ignore ).as_dict() # type: ignore
) # type: ignore ) # type: ignore
elif config.marshoai_enable_support_image_tip: elif config.marshoai_enable_support_image_tip:
await UniMessage("*此模型不支持图片处理。").send() await UniMessage(
"*此模型不支持图片处理或管理员未启用此模型的图片支持。图片将被忽略。"
).send()
backup_context = await get_backup_context(target.id, target.private) backup_context = await get_backup_context(target.id, target.private)
if backup_context: if backup_context:
context.set_context( context.set_context(
@@ -299,20 +284,24 @@ async def marsho(
tools_lists = tools.tools_list + list( tools_lists = tools.tools_list + list(
map(lambda v: v.data(), get_function_calls().values()) map(lambda v: v.data(), get_function_calls().values())
) )
response = await make_chat( response = await make_chat_openai(
client=client, client=client,
model_name=model_name, model_name=model_name,
msg=context_msg + [UserMessage(content=usermsg)], # type: ignore msg=context_msg + [UserMessage(content=usermsg).as_dict()], # type: ignore
tools=tools_lists if tools_lists else None, # TODO 临时追加函数,后期优化 tools=tools_lists if tools_lists else None, # TODO 临时追加函数,后期优化
) )
# await UniMessage(str(response)).send() # await UniMessage(str(response)).send()
choice = response.choices[0] choice = response.choices[0]
if choice["finish_reason"] == CompletionsFinishReason.STOPPED: # Sprint(choice)
# 当tool_calls非空时将finish_reason设置为TOOL_CALLS
if choice.message.tool_calls != None and config.marshoai_fix_toolcalls:
choice.finish_reason = CompletionsFinishReason.TOOL_CALLS
if choice.finish_reason == CompletionsFinishReason.STOPPED:
# 当对话成功时将dict的上下文添加到上下文类中 # 当对话成功时将dict的上下文添加到上下文类中
context.append( context.append(
UserMessage(content=usermsg).as_dict(), target.id, target.private # type: ignore UserMessage(content=usermsg).as_dict(), target.id, target.private # type: ignore
) )
context.append(choice.message.as_dict(), target.id, target.private) context.append(choice.message, target.id, target.private)
if [target.id, target.private] not in target_list: if [target.id, target.private] not in target_list:
target_list.append([target.id, target.private]) target_list.append([target.id, target.private])
@@ -323,7 +312,7 @@ async def marsho(
) )
else: else:
await UniMessage(str(choice.message.content)).send(reply_to=True) await UniMessage(str(choice.message.content)).send(reply_to=True)
elif choice["finish_reason"] == CompletionsFinishReason.CONTENT_FILTERED: elif choice.finish_reason == CompletionsFinishReason.CONTENT_FILTERED:
# 对话失败,消息过滤 # 对话失败,消息过滤
@@ -331,76 +320,89 @@ async def marsho(
reply_to=True reply_to=True
) )
return return
elif choice["finish_reason"] == CompletionsFinishReason.TOOL_CALLS: elif choice.finish_reason == CompletionsFinishReason.TOOL_CALLS:
# function call # function call
# 需要获取额外信息,调用函数工具 # 需要获取额外信息,调用函数工具
tool_msg = [] tool_msg = []
while choice.message.tool_calls != None: while choice.message.tool_calls != None:
tool_msg.append( # await UniMessage(str(response)).send()
AssistantMessage(tool_calls=response.choices[0].message.tool_calls) tool_calls = choice.message.tool_calls
) try:
for tool_call in choice.message.tool_calls: if tool_calls[0]["function"]["name"].startswith("$"):
if isinstance( choice.message.tool_calls[0][
tool_call, ChatCompletionsToolCall "type"
): # 循环调用工具直到不需要调用 ] = "builtin_function" # 兼容 moonshot AI 内置函数的临时方案
try: except:
function_args = json.loads(tool_call.function.arguments) pass
except json.JSONDecodeError: tool_msg.append(choice.message)
function_args = json.loads( for tool_call in tool_calls:
tool_call.function.arguments.replace("'", '"') try:
) function_args = json.loads(tool_call.function.arguments)
logger.info( except json.JSONDecodeError:
f"调用函数 {tool_call.function.name.replace('-', '.')}\n参数:" function_args = json.loads(
+ "\n".join([f"{k}={v}" for k, v in function_args.items()]) tool_call.function.arguments.replace("'", '"')
) )
await UniMessage( # 删除args的placeholder参数
f"调用函数 {tool_call.function.name.replace('-', '.')}\n参数:" if "placeholder" in function_args:
+ "\n".join([f"{k}={v}" for k, v in function_args.items()]) del function_args["placeholder"]
).send() logger.info(
# TODO 临时追加插件函数,若工具中没有则调用插件函数 f"调用函数 {tool_call.function.name.replace('-', '.')}\n参数:"
if tools.has_function(tool_call.function.name): + "\n".join([f"{k}={v}" for k, v in function_args.items()])
logger.debug(f"调用工具函数 {tool_call.function.name}") )
func_return = await tools.call( await UniMessage(
tool_call.function.name, function_args f"调用函数 {tool_call.function.name.replace('-', '.')}\n参数:"
) # 获取返回值 + "\n".join([f"{k}={v}" for k, v in function_args.items()])
else: ).send()
if caller := get_function_calls().get( # TODO 临时追加插件函数,若工具中没有则调用插件函数
tool_call.function.name.replace("-", ".") if tools.has_function(tool_call.function.name):
): logger.debug(f"调用工具函数 {tool_call.function.name}")
logger.debug(f"调用插件函数 {caller.full_name}") func_return = await tools.call(
# 权限检查,规则检查 TODO tool_call.function.name, function_args
# 实现依赖注入检查函数参数及参数注解类型对Event类型的参数进行注入 ) # 获取返回值
func_return = await caller.with_ctx( else:
SessionContext( if caller := get_function_calls().get(tool_call.function.name):
bot=bot, logger.debug(f"调用插件函数 {caller.full_name}")
event=event, # 权限检查,规则检查 TODO
state=state, # 实现依赖注入检查函数参数及参数注解类型对Event类型的参数进行注入
matcher=matcher, func_return = await caller.with_ctx(
) SessionContext(
).call(**function_args) bot=bot,
else: event=event,
logger.error( state=state,
f"未找到函数 {tool_call.function.name.replace('-', '.')}" matcher=matcher,
) )
func_return = f"未找到函数 {tool_call.function.name.replace('-', '.')}" ).call(**function_args)
tool_msg.append( else:
ToolMessage(tool_call_id=tool_call.id, content=func_return) # type: ignore logger.error(
) f"未找到函数 {tool_call.function.name.replace('-', '.')}"
response = await make_chat( )
func_return = f"未找到函数 {tool_call.function.name.replace('-', '.')}"
tool_msg.append(
ToolMessage(tool_call_id=tool_call.id, content=func_return).as_dict() # type: ignore
)
# tool_msg[0]["tool_calls"][0]["type"] = "builtin_function"
# await UniMessage(str(tool_msg)).send()
request_msg = context_msg + [UserMessage(content=usermsg).as_dict()] + tool_msg # type: ignore
response = await make_chat_openai(
client=client, client=client,
model_name=model_name, model_name=model_name,
msg=context_msg + [UserMessage(content=usermsg)] + tool_msg, # type: ignore msg=request_msg, # type: ignore
tools=tools.get_tools_list(), tools=(
tools_lists if tools_lists else None
), # TODO 临时追加函数,后期优化
) )
choice = response.choices[0] choice = response.choices[0]
if choice["finish_reason"] == CompletionsFinishReason.STOPPED: # 当tool_calls非空时将finish_reason设置为TOOL_CALLS
if choice.message.tool_calls != None:
choice.finish_reason = CompletionsFinishReason.TOOL_CALLS
if choice.finish_reason == CompletionsFinishReason.STOPPED:
# 对话成功 添加上下文 # 对话成功 添加上下文
context.append( context.append(
UserMessage(content=usermsg).as_dict(), target.id, target.private # type: ignore UserMessage(content=usermsg).as_dict(), target.id, target.private # type: ignore
) )
# context.append(tool_msg, target.id, target.private) # context.append(tool_msg, target.id, target.private)
context.append(choice.message.as_dict(), target.id, target.private) context.append(choice.message, target.id, target.private)
# 发送消息 # 发送消息
if config.marshoai_enable_richtext_parse: if config.marshoai_enable_richtext_parse:
@@ -410,9 +412,9 @@ async def marsho(
else: else:
await UniMessage(str(choice.message.content)).send(reply_to=True) await UniMessage(str(choice.message.content)).send(reply_to=True)
else: else:
await marsho_cmd.finish(f"意外的完成原因:{choice['finish_reason']}") await marsho_cmd.finish(f"意外的完成原因:{choice.finish_reason}")
else: else:
await marsho_cmd.finish(f"意外的完成原因:{choice['finish_reason']}") await marsho_cmd.finish(f"意外的完成原因:{choice.finish_reason}")
except Exception as e: except Exception as e:
await UniMessage(str(e) + suggest_solution(str(e))).send() await UniMessage(str(e) + suggest_solution(str(e))).send()
traceback.print_exc() traceback.print_exc()
@@ -422,7 +424,7 @@ async def marsho(
with contextlib.suppress(ImportError): # 优化先不做() with contextlib.suppress(ImportError): # 优化先不做()
import nonebot.adapters.onebot.v11 # type: ignore import nonebot.adapters.onebot.v11 # type: ignore
from .azure_onebot import poke_notify from .marsho_onebot import poke_notify
@poke_notify.handle() @poke_notify.handle()
async def poke(event: Event): async def poke(event: Event):
@@ -443,7 +445,7 @@ with contextlib.suppress(ImportError): # 优化先不做()
], ],
) )
choice = response.choices[0] choice = response.choices[0]
if choice["finish_reason"] == CompletionsFinishReason.STOPPED: if choice.finish_reason == CompletionsFinishReason.STOPPED:
await UniMessage(" " + str(choice.message.content)).send( await UniMessage(" " + str(choice.message.content)).send(
at_sender=True at_sender=True
) )
@@ -451,18 +453,3 @@ with contextlib.suppress(ImportError): # 优化先不做()
await UniMessage(str(e) + suggest_solution(str(e))).send() await UniMessage(str(e) + suggest_solution(str(e))).send()
traceback.print_exc() traceback.print_exc()
return return
@driver.on_shutdown
async def auto_backup_context():
for target_info in target_list:
target_id, target_private = target_info
contexts_data = context.build(target_id, target_private)
if target_private:
target_uid = "private_" + target_id
else:
target_uid = "group_" + target_id
await save_context_to_json(
f"back_up_context_{target_uid}", contexts_data, "contexts/backup"
)
logger.info(f"已保存会话 {target_id} 的上下文备份,将在下次对话时恢复~")

View File

@@ -28,40 +28,31 @@ class MarshoContext:
往上下文中添加消息 往上下文中添加消息
""" """
target_dict = self._get_target_dict(is_private) target_dict = self._get_target_dict(is_private)
if target_id not in target_dict: target_dict.setdefault(target_id, []).append(content)
target_dict[target_id] = []
target_dict[target_id].append(content)
def set_context(self, contexts, target_id: str, is_private: bool): def set_context(self, contexts, target_id: str, is_private: bool):
""" """
设置上下文 设置上下文
""" """
target_dict = self._get_target_dict(is_private) self._get_target_dict(is_private)[target_id] = contexts
target_dict[target_id] = contexts
def reset(self, target_id: str, is_private: bool): def reset(self, target_id: str, is_private: bool):
""" """
重置上下文 重置上下文
""" """
target_dict = self._get_target_dict(is_private) self._get_target_dict(is_private).pop(target_id, None)
if target_id in target_dict:
target_dict[target_id].clear()
def reset_all(self): def reset_all(self):
""" """
重置所有上下文 重置所有上下文
""" """
self.contents["private"].clear() self.contents = {"private": {}, "non-private": {}}
self.contents["non-private"].clear()
def build(self, target_id: str, is_private: bool) -> list: def build(self, target_id: str, is_private: bool) -> list:
""" """
构建返回的上下文,不包括系统消息 构建返回的上下文,不包括系统消息
""" """
target_dict = self._get_target_dict(is_private) return self._get_target_dict(is_private).setdefault(target_id, [])
if target_id not in target_dict:
target_dict[target_id] = []
return target_dict[target_id]
class MarshoTools: class MarshoTools:
@@ -84,54 +75,52 @@ class MarshoTools:
for package_name in os.listdir(tools_dir): for package_name in os.listdir(tools_dir):
package_path = os.path.join(tools_dir, package_name) package_path = os.path.join(tools_dir, package_name)
# logger.info(f"尝试加载工具包 {package_name}")
if package_name in config.marshoai_disabled_toolkits: if package_name in config.marshoai_disabled_toolkits:
logger.info(f"工具包 {package_name} 已被禁用。") logger.info(f"工具包 {package_name} 已被禁用。")
continue continue
if os.path.isdir(package_path) and os.path.exists( if os.path.isdir(package_path) and os.path.exists(
os.path.join(package_path, "__init__.py") os.path.join(package_path, "__init__.py")
): ):
json_path = os.path.join(package_path, "tools.json") self._load_package(package_name, package_path)
if os.path.exists(json_path):
try:
with open(json_path, "r", encoding="utf-8") as json_file:
data = json.load(json_file)
for i in data:
self.tools_list.append(i)
spec = importlib.util.spec_from_file_location(
package_name, os.path.join(package_path, "__init__.py")
)
package = importlib.util.module_from_spec(spec)
self.imported_packages[package_name] = package
sys.modules[package_name] = package
spec.loader.exec_module(package)
logger.success(f"成功加载工具包 {package_name}")
except json.JSONDecodeError as e:
logger.error(f"解码 JSON {json_path} 时发生错误: {e}")
except Exception as e:
logger.error(f"加载工具包时发生错误: {e}")
traceback.print_exc()
else:
logger.warning(
f"在工具包 {package_path} 下找不到tools.json跳过加载。"
)
else: else:
logger.warning(f"{package_path} 不是有效的工具包路径,跳过加载。") logger.warning(f"{package_path} 不是有效的工具包路径,跳过加载。")
def _load_package(self, package_name, package_path):
json_path = os.path.join(package_path, "tools.json")
if os.path.exists(json_path):
try:
with open(json_path, "r", encoding="utf-8") as json_file:
data = json.load(json_file)
self.tools_list.extend(data)
spec = importlib.util.spec_from_file_location(
package_name, os.path.join(package_path, "__init__.py")
)
package = importlib.util.module_from_spec(spec)
self.imported_packages[package_name] = package
sys.modules[package_name] = package
spec.loader.exec_module(package)
logger.success(f"成功加载工具包 {package_name}")
except json.JSONDecodeError as e:
logger.error(f"解码 JSON {json_path} 时发生错误: {e}")
except Exception as e:
logger.error(f"加载工具包时发生错误: {e}")
traceback.print_exc()
else:
logger.warning(f"在工具包 {package_path} 下找不到tools.json跳过加载。")
async def call(self, full_function_name: str, args: dict): async def call(self, full_function_name: str, args: dict):
""" """
调用指定的函数 调用指定的函数
""" """
# 分割包名和函数名
parts = full_function_name.split("__") parts = full_function_name.split("__")
if len(parts) == 2: if len(parts) != 2:
package_name = parts[0]
function_name = parts[1]
else:
logger.error("函数名无效") logger.error("函数名无效")
return
package_name, function_name = parts
if package_name in self.imported_packages: if package_name in self.imported_packages:
package = self.imported_packages[package_name] package = self.imported_packages[package_name]
try: try:
@@ -149,12 +138,11 @@ class MarshoTools:
检查是否存在指定的函数 检查是否存在指定的函数
""" """
try: try:
for t in self.tools_list: return any(
if t["function"]["name"].replace( t["function"]["name"].replace("-", "_")
"-", "_" == full_function_name.replace("-", "_")
) == full_function_name.replace("-", "_"): for t in self.tools_list
return True )
return False
except Exception as e: except Exception as e:
logger.error(f"检查函数 '{full_function_name}' 时发生错误:{e}") logger.error(f"检查函数 '{full_function_name}' 时发生错误:{e}")
return False return False

View File

@@ -17,11 +17,21 @@ _caller_data: dict[str, "Caller"] = {}
class Caller: class Caller:
def __init__(self, name: str = "", description: str | None = None): def __init__(
self,
name: str = "",
description: str | None = None,
func_type: str = "function",
no_module_name: bool = False,
):
self._name: str = name self._name: str = name
"""函数名称""" """函数名称"""
self._description = description self._description = description
"""函数描述""" """函数描述"""
self._func_type = func_type
"""函数类型"""
self.no_module_name = no_module_name
"""是否不包含模块名"""
self._plugin: Plugin | None = None self._plugin: Plugin | None = None
"""所属插件对象,装饰时声明""" """所属插件对象,装饰时声明"""
self.func: ASYNC_FUNCTION_CALL_FUNC | None = None self.func: ASYNC_FUNCTION_CALL_FUNC | None = None
@@ -142,7 +152,7 @@ class Caller:
module_name = "" module_name = ""
self.module_name = module_name self.module_name = module_name
_caller_data[self.full_name] = self _caller_data[self.aifc_name] = self
logger.opt(colors=True).debug( logger.opt(colors=True).debug(
f"<y>加载函数 {self.full_name}: {self._description}</y>" f"<y>加载函数 {self.full_name}: {self._description}</y>"
) )
@@ -155,16 +165,20 @@ class Caller:
Returns: Returns:
dict[str, Any]: 函数的json数据 dict[str, Any]: 函数的json数据
""" """
properties = {key: value.data() for key, value in self._parameters.items()}
if not properties:
properties["placeholder"] = {
"type": "string",
"description": "占位符,用于显示在对话框中", # 为保证兼容性而设置的无用参数
}
return { return {
"type": "function", "type": self._func_type,
"function": { "function": {
"name": self.aifc_name, "name": self.aifc_name,
"description": self._description, "description": self._description,
"parameters": { "parameters": {
"type": "object", "type": "object",
"properties": { "properties": properties,
key: value.data() for key, value in self._parameters.items()
},
}, },
"required": [ "required": [
key key
@@ -233,7 +247,9 @@ class Caller:
@property @property
def aifc_name(self) -> str: def aifc_name(self) -> str:
"""AI调用名没有点""" """AI调用名没有点"""
return self._name.replace(".", "-") if self.no_module_name:
return self._name
return self.full_name.replace(".", "-")
@property @property
def full_name(self) -> str: def full_name(self) -> str:
@@ -245,16 +261,29 @@ class Caller:
return f"{self.full_name}({self._description})" return f"{self.full_name}({self._description})"
def on_function_call(name: str = "", description: str | None = None) -> Caller: def on_function_call(
name: str = "",
description: str | None = None,
func_type: str = "function",
no_module_name: bool = False,
) -> Caller:
"""返回一个Caller类可用于装饰一个函数使其注册为一个可被AI调用的function call函数 """返回一个Caller类可用于装饰一个函数使其注册为一个可被AI调用的function call函数
Args: Args:
name: 函数名称若为空则从函数的__name__属性获取
description: 函数描述若为None则从函数的docstring中获取 description: 函数描述若为None则从函数的docstring中获取
func_type: 函数类型默认为function若要注册为 Moonshot AI 的内置函数则为builtin_function
no_module_name: 是否不包含模块名,当注册为 Moonshot AI 的内置函数时为True
Returns: Returns:
Caller: Caller对象 Caller: Caller对象
""" """
caller = Caller(name=name, description=description) caller = Caller(
name=name,
description=description,
func_type=func_type,
no_module_name=no_module_name,
)
return caller return caller

View File

@@ -107,9 +107,7 @@ def load_plugins(*plugin_dirs: str) -> set[Plugin]:
for plugin_dir in plugin_dirs: for plugin_dir in plugin_dirs:
for f in os.listdir(plugin_dir): for f in os.listdir(plugin_dir):
path = Path(os.path.join(plugin_dir, f)) path = Path(os.path.join(plugin_dir, f))
module_name = None module_name = None
if os.path.isfile(path) and f.endswith(".py"): if os.path.isfile(path) and f.endswith(".py"):
"""单文件加载""" """单文件加载"""
module_name = f"{path_to_module_name(Path(plugin_dir))}.{f[:-3]}" module_name = f"{path_to_module_name(Path(plugin_dir))}.{f[:-3]}"

View File

@@ -1,6 +1,6 @@
from nonebot_plugin_marshoai.plugin import PluginMetadata from nonebot_plugin_marshoai.plugin import PluginMetadata
from .chat import * # from .chat import *
from .file_io import * from .file_io import *
from .liteyuki import * from .liteyuki import *
from .manager import * from .manager import *

View File

@@ -1,7 +1,7 @@
import traceback import traceback
import httpx import httpx
from zhDateTime import DateTime from zhDateTime import DateTime # type: ignore
from nonebot_plugin_marshoai.plugin import PluginMetadata, on_function_call from nonebot_plugin_marshoai.plugin import PluginMetadata, on_function_call
from nonebot_plugin_marshoai.plugin.func_call.params import String from nonebot_plugin_marshoai.plugin.func_call.params import String
@@ -49,7 +49,7 @@ async def get_bangumi_news() -> str:
for item in items: for item in items:
name = item["name_cn"] name = item["name_cn"]
info += f"{name}" info += f"{name}"
info += "" info += "\n"
return info return info
except Exception as e: except Exception as e:
traceback.print_exc() traceback.print_exc()

View File

@@ -0,0 +1,45 @@
from nonebot_plugin_marshoai.plugin import (
Integer,
Parameter,
PluginMetadata,
String,
on_function_call,
)
from . import mk_morse_code, mk_nya_code
__marsho_meta__ = PluginMetadata(
name="MegaKits插件",
description="一个功能混杂的多文件插件",
author="Twisuki",
)
@on_function_call(description="摩尔斯电码加密").params(
msg=String(description="被加密语句")
)
async def morse_encrypt(msg: str) -> str:
"""摩尔斯电码加密"""
return str(await mk_morse_code.morse_encrypt(msg))
@on_function_call(description="摩尔斯电码解密").params(
msg=String(description="被解密语句")
)
async def morse_decrypt(msg: str) -> str:
"""摩尔斯电码解密"""
return str(await mk_morse_code.morse_decrypt(msg))
@on_function_call(description="转换为猫语").params(msg=String(description="被转换语句"))
async def nya_encrypt(msg: str) -> str:
"""转换为猫语"""
return str(await mk_nya_code.nya_encrypt(msg))
@on_function_call(description="将猫语翻译回人类语言").params(
msg=String(description="被翻译语句")
)
async def nya_decrypt(msg: str) -> str:
"""将猫语翻译回人类语言"""
return str(await mk_nya_code.nya_decrypt(msg))

View File

@@ -0,0 +1,82 @@
# MorseCode
MorseEncode = {
"A": ".-",
"B": "-...",
"C": "-.-.",
"D": "-..",
"E": ".",
"F": "..-.",
"G": "--.",
"H": "....",
"I": "..",
"J": ".---",
"K": "-.-",
"L": ".-..",
"M": "--",
"N": "-.",
"O": "---",
"P": ".--.",
"Q": "--.-",
"R": ".-.",
"S": "...",
"T": "-",
"U": "..-",
"V": "...-",
"W": ".--",
"X": "-..-",
"Y": "-.--",
"Z": "--..",
"1": ".----",
"2": "..---",
"3": "...--",
"4": "....-",
"5": ".....",
"6": "-....",
"7": "--...",
"8": "---..",
"9": "----.",
"0": "-----",
".": ".-.-.-",
":": "---...",
",": "--..--",
";": "-.-.-.",
"?": "..--..",
"=": "-...-",
"'": ".----.",
"/": "-..-.",
"!": "-.-.--",
"-": "-....-",
"_": "..--.-",
'"': ".-..-.",
"(": "-.--.",
")": "-.--.-",
"$": "...-..-",
"&": "....",
"@": ".--.-.",
" ": " ",
}
MorseDecode = {value: key for key, value in MorseEncode.items()}
async def morse_encrypt(msg: str):
result = ""
msg = msg.upper()
for char in msg:
if char in MorseEncode:
result += MorseEncode[char]
else:
result += "..--.."
result += " "
return result
async def morse_decrypt(msg: str):
result = ""
msg = msg.replace("_", "-")
msg_arr = msg.split(" ")
for element in msg_arr:
if element in MorseDecode:
result += MorseDecode[element]
else:
result += "?"
return result

View File

@@ -0,0 +1,74 @@
# NyaCode
import base64
import random
NyaCodeCharset = ["", "", "?", "~"]
NyaCodeSpecialCharset = ["", "!", "...", ".."]
NyaCodeEncode = {}
for i in range(64):
triplet = ""
for j in range(3):
index = (i // (4**j)) % 4
triplet += NyaCodeCharset[index]
if i < 26:
char = chr(65 + i) # 大写字母 A-Z
elif i < 52:
char = chr(97 + (i - 26)) # 小写字母 a-z
elif i < 62:
char = chr(48 + (i - 52)) # 数字 0-9
elif i == 62:
char = chr(43) # 特殊字符 +
else:
char = chr(47) # 特殊字符 /
NyaCodeEncode[char] = triplet
NyaCodeDecode = {value: key for key, value in NyaCodeEncode.items()}
async def nya_encrypt(msg: str):
result = ""
b64str = base64.b64encode(msg.encode()).decode().replace("=", "")
nyastr = ""
for b64char in b64str:
nyastr += NyaCodeEncode[b64char]
for char in nyastr:
if char == "" and random.random() < 0.5:
result += "!"
if random.random() < 0.25:
result += random.choice(NyaCodeSpecialCharset) + char
else:
result += char
return result
async def nya_decrypt(msg: str):
msg = msg.replace("", "").replace("!", "").replace(".", "")
nyastr = []
i = 0
if len(msg) % 3 != 0:
return "这句话不是正确的猫语"
while i < len(msg):
nyachar = msg[i : i + 3]
try:
if all(char in NyaCodeCharset for char in nyachar):
nyastr.append(nyachar)
i += 3
except Exception:
return "这句话不是正确的猫语"
b64str = ""
for nyachar in nyastr:
b64str += NyaCodeDecode[nyachar]
b64str += "=" * (4 - len(b64str) % 4)
try:
result = base64.b64decode(b64str.encode()).decode()
except Exception:
return "翻译失败"
return result

View File

@@ -0,0 +1,79 @@
from nonebot_plugin_marshoai.plugin import (
Integer,
Parameter,
PluginMetadata,
String,
on_function_call,
)
from . import pc_cat, pc_info, pc_shop, pc_token
__marsho_meta__ = PluginMetadata(
name="养猫插件",
description="在Marsho这里赛博养猫",
author="Twisuki",
)
# 交互
@on_function_call(description="传入猫猫种类, 新建一只猫猫").params(
type=String(description='猫猫种类, 默认"猫1", 可留空')
)
async def cat_new(type: str) -> str:
"""新建猫猫"""
return pc_cat.cat_new(type)
@on_function_call(
description="传入token(一串长20的b64字符串), 新名字, 选用技能, 进行猫猫的初始化"
).params(
token=String(description="token(一串长20的b64字符串)"),
name=String(description="新名字"),
skill=String(description="技能"),
)
async def cat_init(token: str, name: str, skill: str) -> str:
"""初始化猫猫"""
return pc_cat.cat_init(token, name, skill)
@on_function_call(description="传入token, 查看猫猫信息").params(
token=String(description="token(一串长20的b64字符串)"),
)
async def cat_show(token: str) -> str:
"""查询信息"""
return pc_cat.cat_show(token)
@on_function_call(description="传入token, 玩猫").params(
token=String(description="token(一串长20的b64字符串)"),
)
async def cat_play(token: str) -> str:
"""玩猫"""
return pc_cat.cat_play(token)
@on_function_call(description="传入token, 投喂猫猫").params(
token=String(description="token(一串长20的b64字符串)"),
)
async def cat_feed(token: str) -> str:
"""喂猫"""
return pc_cat.cat_feed(token)
# 帮助
@on_function_call(description="帮助文档/如何创建一只猫猫").params()
async def help_cat_new() -> str:
return pc_info.help_cat_new()
@on_function_call(description="可选种类").params()
async def help_cat_type() -> str:
return pc_info.print_type_list()
@on_function_call(description="可选技能").params()
async def help_cat_skill() -> str:
return pc_info.print_skill_list()
# 商店

View File

@@ -0,0 +1,241 @@
# 主交互
import functools
from datetime import datetime
from typing import List
from nonebot.log import logger
from . import pc_info, pc_token
from .pc_info import SKILL_LIST, TYPE_LIST, value_output
from .pc_token import dict_to_token, token_to_dict
"""特判标准
1. 默认, 未初始化: name = Default0
2. 错误: name = ERROR!
3. 死亡: skill = [False] * 8
"""
# 私用列表
DEFAULT_DICT = {
"name": "Default0",
"age": 0,
"type": 0,
"health": 0,
"saturation": 0,
"energy": 0,
"skill": [False] * 8,
"date": 0,
}
DEFAULT_TOKEN = "6IyszC6tjoYAAAAAAAAC"
# 交互前数据更新
def cat_update(func):
@functools.wraps(func)
def wrapper(*args, **kwargs):
if args:
token = args[0]
data = token_to_dict(token)
# 检查
if data["name"] == "Default0":
return "猫猫尚未初始化, 请初始化猫猫"
if data["name"] == "ERROR!":
return (
"token出错"
f'token应为Base64字符串, 当前token : "{token}"'
f"当前token长度应为20, 当前长度 : {len(token)}"
)
if data["skill"] == [False] * 8:
return (
"很不幸, 猫猫已死亡"
f'名字 : {data["name"]}'
f'年龄 : {data["age"]}'
)
date = data["date"]
now = (datetime(2025, 1, 1) - datetime.now()).days
# 喂食状态更新
if now - date > 5:
data["saturation"] = max(data["saturation"] - 64, 0)
data["health"] = max(data["health"] - 32, 0)
data["energy"] = max(data["energy"] - 32, 0)
elif now - date > 2:
data["saturation"] = max(data["saturation"] - 16, 0)
data["health"] = max(data["health"] - 8, 0)
data["energy"] = max(data["energy"] - 16, 0)
# 机能状态更新
if data["saturation"] / 1.27 < 20:
data["health"] = max(data["health"] - 8, 0)
elif data["saturation"] / 1.27 > 80:
data["health"] = min(data["health"] + 8, 127)
# 生长检查
if now % 7 == 0:
# 死亡
if data["health"] / 1.27 < 20:
data["health"] = 0
death = DEFAULT_DICT
death["name"] = data["name"]
data = death
# 生长
if data["health"] / 1.27 > 60 and data["saturation"] / 1.27 > 40:
data["age"] = min(data["age"] + 1, 15)
token = dict_to_token(data)
new_args = (token,) + args[1:]
return func(*new_args, **kwargs)
return wrapper
# 创建对象
def cat_new(type: str = "猫1") -> str:
data = DEFAULT_DICT
if type not in TYPE_LIST:
return (
f'未知的"{type}"种类, 请重新选择.'
f"\n可选种类 : {pc_info.print_type_list()}"
)
data["type"] = TYPE_LIST.index(type)
token = dict_to_token(data)
return (
f'猫猫已创建, 种类为 : "{type}"; \ntoken : "{token}",'
f"\n请妥善保存token, 这是猫猫的唯一标识符!"
f"\n新的猫猫还没有起名字, 请对猫猫进行初始化, 起一个长度小于等于8位的名字(仅限大小写字母+数字+特殊符号), 并选取一个技能."
f"\n技能列表 : {pc_info.print_skill_list()}"
)
# 初始化对象
def cat_init(token: str, name: str, skill: str) -> str:
data = token_to_dict(token)
if data["name"] != "Default0":
logger.info("初始化失败!")
return "该猫猫已进行交互, 无法进行初始化!"
if skill not in SKILL_LIST:
return (
f'未知的"{skill}"技能, 请重新选择.'
f"技能列表 : {pc_info.print_skill_list()}"
)
data["name"] = name
data["skill"][SKILL_LIST.index(skill)] = True
data["health"] = 127
data["saturation"] = 127
data["energy"] = 127
token = dict_to_token(data)
return (
f'初始化完成, 名字 : "{data["name"]}", 种类 : "{data["type"]}", 技能 : "{skill}"'
f'\n新token : "{token}"'
f"\n请妥善保存token, 这是猫猫的唯一标识符!"
)
# 查看信息
@cat_update
def cat_show(token: str) -> str:
result = pc_info.print_info(token)
data = token_to_dict(token)
if data["health"] / 1.27 < 20:
return result + "\n猫猫健康状况非常差! 甚至濒临死亡!! 请立即前往医院救治!!"
if data["health"] / 1.27 < 60:
result += "\n猫猫健康状况较差, 请投喂食物或陪猫猫玩耍"
if data["saturation"] / 1.27 < 40:
result += "\n猫猫很饿, 请投喂食物"
if data["energy"] / 1.27 < 20:
result += "\n猫猫很累, 请抱猫睡觉, 不要投喂食物或陪它玩耍"
return result
# 陪猫猫玩耍
@cat_update
def cat_play(token: str) -> str:
data = token_to_dict(token)
if data["health"] / 1.27 < 20:
return "猫猫健康状况非常差! 甚至濒临死亡!! 请立即前往医院救治!!"
if data["saturation"] / 1.27 < 40:
return "猫猫很饿, 拒接玩耍请求."
if data["energy"] / 1.27 < 20:
return "猫猫很累, 拒接玩耍请求"
data["health"] = min(data["health"] + 16, 127)
data["saturation"] = max(data["saturation"] - 16, 0)
data["energy"] = max(data["energy"] - 8, 0)
token = dict_to_token(data)
return (
f'你陪猫猫玩耍了一个小时, 猫猫的生命值上涨到了{value_output(data["health"])}'
f'\n新token : "{token}"'
"\n请妥善保存token, 这是猫猫的唯一标识符!"
)
# 喂食
@cat_update
def cat_feed(token: str) -> str:
data = token_to_dict(token)
if data["health"] / 1.27 < 20:
return "猫猫健康状况非常差! 甚至濒临死亡!! 请立即前往医院救治!!"
if data["saturation"] / 1.27 > 80:
return "猫猫并不饿, 不需要喂食"
if data["energy"] / 1.27 < 40:
return "猫猫很累, 请抱猫睡觉, 不要投喂食物或陪它玩耍"
data["saturation"] = min(data["saturation"] + 32, 127)
data["date"] = (datetime(2025, 1, 1) - datetime.now()).days
token = dict_to_token(data)
return (
f'你投喂了2单位标准猫粮, 猫猫的饱食度提升到了{value_output(data["saturation"])}'
f'\n新token : "{token}"'
"\n请妥善保存token, 这是猫猫的唯一标识符!"
)
# 睡觉
@cat_update
def cat_sleep(token: str) -> str:
data = token_to_dict(token)
if data["health"] / 1.27 < 20:
return "猫猫健康状况非常差! 甚至濒临死亡!! 请立即前往医院救治!!"
if data["saturation"] / 1.27 < 40:
return "猫猫很饿, 请喂食."
if data["energy"] / 1.27 > 80:
return "猫猫很精神, 不需要睡觉"
data["health"] = min(data["health"] + 8, 127)
data["energy"] = min(data["energy"] + 16, 0)
token = dict_to_token(data)
return (
f'你抱猫休息了一阵子, 猫猫的活力值提升到了{value_output(data["energy"])}'
f'\n新token : "{token}"'
"\n请妥善保存token, 这是猫猫的唯一标识符!"
)
"""
1. 商店系统
2. 技能系统
3. 提高复杂性和难度
"""

View File

@@ -0,0 +1,77 @@
# 插件使用复杂, 这里用作输出提示信息.
# 如: 帮助, 每次操作后对猫猫状态的描述\打印特殊列表
# 公用列表数据转到这里存储
from nonebot.log import logger
from .pc_token import dict_to_token, token_to_dict
# 公用列表
TYPE_LIST = ["猫1", "猫2", "猫3", "猫4", "猫5", "猫6", "猫7", "猫8"]
SKILL_LIST = ["s1", "s2", "s3", "s4", "s5", "s6", "s7", "s8"]
# 提示词打印
# 打印种类列表
def print_type_list() -> str:
result = ""
for type in TYPE_LIST:
result += f'"{type}", '
result = result[:-2]
return f"({result})"
# 打印技能列表
def print_skill_list() -> str:
result = ""
for skill in SKILL_LIST:
result += f'"{skill}", '
result = result[:-2]
return f"({result})"
# 127位值 - 100%快速转换
def value_output(num: int) -> str:
value = int(num / 1.27)
return str(value)
# 打印状态
def print_info(token: str) -> str:
data = token_to_dict(token)
return (
"状态信息: "
f'\n\t名字 : {data["name"]}'
f'\n\t种类 : {TYPE_LIST[data["type"]]}'
f'\n\t生命值 : {value_output(data["health"])}'
f'\n\t饱食度 : {value_output(data["saturation"])}'
f"\n\t活力值 : {value_output(data['energy'])}"
f"\n\t技能 : {print_skill(token)}"
f"\n新token : {token}"
f"\ntoken已更新, 请妥善保存token, 这是猫猫的唯一标识符!"
)
# 打印已有技能
def print_skill(token: str) -> str:
result = ""
data = token_to_dict(token)
for index in range(0, len(SKILL_LIST) - 1):
if data["skill"][index]:
result += f"{SKILL_LIST[index]}, "
logger.info(data["skill"])
return result[:-2]
# 帮助
# 创建猫猫
def help_cat_new() -> str:
return (
"新建一只猫猫, 首先选择猫猫的种类, 获取初始化token;"
"然后用这个token, 选择名字和一个技能进行初始化;"
"初始化结束才表示猫猫正式创建成功."
"\ntoken为猫的唯一标识符, 每次交互都需要传入token"
f"\n种类可选 : {print_type_list()}"
f"\n技能可选 : {print_skill_list()}"
)

View File

@@ -0,0 +1 @@
# 商店

View File

@@ -0,0 +1,200 @@
# 由于无法直接存储数据, 使用一个字符串记录全部信息
# 这里作为引用进行编码/解码, 以及公用数据
"""猫对象属性存储编码Token
名字: 3位长度 + 8位ASCII字符 - 67b
年龄: 0 ~ 15 - 4b
种类: 8种 - 3b
生命值: 0 ~ 127 - 7b
饱食度: 0 ~ 127 - 7b
活力值: 0 ~ 127 - 7b
技能: 8种任选 - 8b
时间: 0 ~ 131017d > 2025-1-1 - 17b
总计120b有效数据
总计120b数据, 15字节, 每3字节(utf-8一个字符)转换为4个Base64字符
总计20个Base64字符的字符串
"""
"""定义变量
存储字符串: Token: str
对象数据: data: dict
名字: name: str
年龄: age: int
种类: type: int # 代表TYPE_LIST中的序号
生命值: health: int # 0 - 127, 显示时转换为100%
饱食度: saturation: int # 0 - 127, 显示时转换为100%
活力值: energy: int # 0 - 127, 显示时转换为100%
技能: skill: List[bool] # 8元素bool数组
时间: date: int # 到2025-1-1按日的时间戳
"""
import base64
from datetime import datetime
from typing import List
from nonebot.log import logger
# 私用列表
ERROR_DICT = {
"name": "ERROR!",
"age": 0,
"type": 0,
"health": 0,
"saturation": 0,
"energy": 0,
"skill": [False, False, False, False, False, False, False, False],
"date": 0,
}
ERROR_TOKEN = "yKpKSepEIAAAAAAAAAAA"
# bool数组/int数据转换
def bool_to_int(bool_array: List[bool]) -> int:
result = 0
for index, bit in enumerate(bool_array[::-1]):
if bit:
result |= 1 << index
return result
def int_to_bool(integer: int, length: int = 0) -> List[bool]:
bit_length = integer.bit_length()
bool_array = [False] * bit_length
for i in range(bit_length):
if integer & (1 << i):
bool_array[bit_length - 1 - i] = True
if len(bool_array) >= length:
return bool_array
else:
return [*([False] * (length - len(bool_array))), *bool_array]
# bool数组/byte数据转换
def bool_to_byte(bool_array: List[bool]) -> bytes:
byte_data = bytearray()
for i in range(0, len(bool_array), 8):
byte = 0
for j in range(8):
if i + j < len(bool_array) and bool_array[i + j]:
byte |= 1 << (7 - j)
byte_data.append(byte)
return bytes(byte_data)
def byte_to_bool(byte_data: bytes, length: int = 0) -> List[bool]:
bool_array = []
for byte in byte_data:
for bit in format(byte, "08b"):
bool_array.append(bit == "1")
if len(bool_array) >= length:
return bool_array
else:
return [*([False] * (length - len(bool_array))), *bool_array]
# 数据解码
def token_to_dict(token: str) -> dict:
logger.info(f"开始解码...\n{token}")
data = {
"name": "Default0",
"age": 0,
"type": 0,
"health": 0,
"saturation": 0,
"energy": 0,
"skill": [False] * 8,
"date": 0,
}
# 转换token
try:
token_byte = base64.b64decode(token.encode())
code = byte_to_bool(token_byte)
except ValueError:
logger.error("token b64解码错误!")
return ERROR_DICT
# 拆分code
name_length = bool_to_int(code[0:3]) + 1
name_code = code[3:67]
age = bool_to_int(code[67:71])
type = bool_to_int(code[71:74])
health = bool_to_int(code[74:81])
saturation = bool_to_int(code[81:88])
energy = bool_to_int(code[88:95])
skill = code[95:103]
date = bool_to_int(code[103:120])
# 解析code
name: str = ""
try:
for i in range(name_length):
character_code = bool_to_byte(name_code[8 * i : 8 * i + 8])
name += character_code.decode("ASCII")
except UnicodeDecodeError:
logger.error("token ASCII解析错误!")
return ERROR_DICT
data["name"] = name
data["age"] = age
data["type"] = type
data["health"] = health
data["saturation"] = saturation
data["energy"] = energy
data["skill"] = skill
data["date"] = date
logger.success(f"解码完成, 数据为\n{data}")
return data
# 数据编码
def dict_to_token(data: dict) -> str:
logger.info(f"开始编码...\n{data}")
code = [False] * 120
# 拆分data
name_length = len(data["name"])
if name_length > 8:
logger.error("name过长")
return ERROR_TOKEN
name = data["name"]
age = data["age"]
type = data["type"]
health = data["health"]
saturation = data["saturation"]
energy = data["energy"]
skill = data["skill"]
date = data["date"]
# 填入code
code[0:3] = int_to_bool(name_length - 1, 3)
name_code = [False] * 64
try:
for i in range(name_length):
character_code = byte_to_bool(name[i].encode("ASCII"), 8)
name_code[8 * i : 8 * i + 8] = character_code
except UnicodeEncodeError:
# "name": "ERROR!"
logger.error("name内含有非法字符!")
return ERROR_TOKEN
code[3:67] = name_code
code[67:71] = int_to_bool(age, 4)
code[71:74] = int_to_bool(type, 3)
code[74:81] = int_to_bool(health, 7)
code[81:88] = int_to_bool(saturation, 7)
code[88:95] = int_to_bool(energy, 7)
code[95:103] = skill
code[103:120] = int_to_bool(date, 17)
# 转换token
token_byte = bool_to_byte(code)
token = base64.b64encode(token_byte).decode()
logger.success(f"编码完成, token为\n{token}")
return token

View File

@@ -1,8 +1,16 @@
import os import os
from zhDateTime import DateTime from zhDateTime import DateTime # type: ignore
from nonebot_plugin_marshoai.plugin import String, on_function_call from nonebot_plugin_marshoai.plugin import PluginMetadata, String, on_function_call
# from .web import *
# 定义插件元数据
__marsho_meta__ = PluginMetadata(
name="基本功能",
author="MarshoAI",
description="这个插件提供基本的功能,比如获取当前时间和日期。",
)
@on_function_call(description="获取当前时间,日期和星期") @on_function_call(description="获取当前时间,日期和星期")

View File

@@ -0,0 +1,97 @@
import json
from pathlib import Path
from azure.ai.inference.models import UserMessage
from nonebot import get_driver, logger, require
from nonebot_plugin_localstore import get_plugin_data_file
require("nonebot_plugin_apscheduler")
require("nonebot_plugin_marshoai")
from nonebot_plugin_apscheduler import scheduler
from nonebot_plugin_marshoai.instances import client
from nonebot_plugin_marshoai.plugin import PluginMetadata, on_function_call
from nonebot_plugin_marshoai.plugin.func_call.params import String
from .command import *
from .config import plugin_config
__marsho_meta__ = PluginMetadata(
name="记忆保存",
author="MarshoAI",
description="这个插件可以帮助AI记住一些事情",
)
memory_path = get_plugin_data_file("memory.json")
if not Path(memory_path).exists():
with open(memory_path, "w", encoding="utf-8") as f:
json.dump({}, f, ensure_ascii=False, indent=4)
# print(memory_path)
driver = get_driver()
@on_function_call(
description="当你发现与你对话的用户的一些信息值得你记忆,或者用户让你记忆等时,调用此函数存储记忆内容"
).params(
memory=String(description="你想记住的内容,概括并保留关键内容"),
user_id=String(description="你想记住的人的id"),
)
async def write_memory(memory: str, user_id: str):
with open(memory_path, "r", encoding="utf-8") as f:
memory_data = json.load(f)
memorys = memory_data.get(user_id, [])
memorys.append(memory)
memory_data[user_id] = memorys
with open(memory_path, "w", encoding="utf-8") as f:
json.dump(memory_data, f, ensure_ascii=False, indent=4)
return "记忆已经保存啦~"
@on_function_call(
description="你需要回忆有关用户的一些知识时,调用此函数读取记忆内容,当用户问问题的时候也尽量调用此函数参考"
).params(user_id=String(description="你想读取记忆的人的id"))
async def read_memory(user_id: str):
with open(memory_path, "r", encoding="utf-8") as f:
memory_data = json.load(f)
memorys = memory_data.get(user_id, [])
if not memorys:
return "好像对ta还没有任何记忆呢~"
return "这些是有关ta的记忆" + "\n".join(memorys)
async def organize_memories():
with open(memory_path, "r", encoding="utf-8") as f:
memory_data = json.load(f)
for i in memory_data:
memory_data_ = "\n".join(memory_data[i])
msg = f"这是一些大模型记忆信息,请你保留重要内容,尽量减少无用的记忆后重新输出记忆内容,浓缩为一行:\n{memory_data_}"
res = await client.complete(UserMessage(content=msg))
try:
memory = res.choices[0].message.content # type: ignore
memory_data[i] = memory
except AttributeError:
logger.error(f"整理关于{i}的记忆时出错:{res}")
with open(memory_path, "w", encoding="utf-8") as f:
json.dump(memory_data, f, ensure_ascii=False, indent=4)
if plugin_config.marshoai_plugin_memory_scheduler:
@driver.on_startup
async def _():
logger.info("小棉定时记忆整理已启动!")
scheduler.add_job(
organize_memories,
"cron",
hour="0",
minute="0",
second="0",
day="*",
id="organize_memories",
)

View File

@@ -0,0 +1,47 @@
import json
from arclet.alconna import Alconna, Args, Subcommand
from nonebot import logger
from nonebot.adapters import Bot, Event
from nonebot.matcher import Matcher
from nonebot.typing import T_State
from nonebot_plugin_alconna import on_alconna
from nonebot_plugin_localstore import get_plugin_data_file
from nonebot_plugin_marshoai.config import config
marsho_memory_cmd = on_alconna(
Alconna(
f"{config.marshoai_default_name}.memory",
Subcommand("view", alias={"v"}),
Subcommand("reset", alias={"r"}),
),
priority=10,
block=True,
)
memory_path = get_plugin_data_file("memory.json")
@marsho_memory_cmd.assign("view")
async def view_memory(matcher: Matcher, state: T_State, event: Event):
user_id = str(event.get_user_id())
with open(memory_path, "r", encoding="utf-8") as f:
memory_data = json.load(f)
memorys = memory_data.get(user_id, [])
if not memorys:
await matcher.finish("好像对ta还没有任何记忆呢~")
await matcher.finish("这些是有关ta的记忆" + "\n".join(memorys))
@marsho_memory_cmd.assign("reset")
async def reset_memory(matcher: Matcher, state: T_State, event: Event):
user_id = str(event.get_user_id())
with open(memory_path, "r", encoding="utf-8") as f:
memory_data = json.load(f)
if user_id in memory_data:
del memory_data[user_id]
with open(memory_path, "w", encoding="utf-8") as f:
json.dump(memory_data, f, ensure_ascii=False, indent=4)
await matcher.finish("记忆已重置~")
await matcher.finish("没有找到该用户的记忆~")

View File

@@ -0,0 +1,9 @@
from nonebot import get_plugin_config, logger
from pydantic import BaseModel
class ConfigModel(BaseModel):
marshoai_plugin_memory_scheduler: bool = True
plugin_config: ConfigModel = get_plugin_config(ConfigModel)

View File

@@ -0,0 +1,17 @@
[
{
"type": "function",
"function": {
"name": "marshoai_memory__write_memory",
"description": "如果在上下中你看见并觉得应该记住的人的行为与事件请调用这个函数并将记忆内容写入。请尽量每次都调用总结ta的习惯、爱好和性格,以及你对ta的印象和ta对你的印象;比如用户喜欢干什么吃什么。"
}
},
{
"type": "function",
"function": {
"name": "marshoai_memory__read_memory",
"description": "每当你想要获取更多有关某人的信息的时候,请调用这个函数。"
}
}
]

View File

@@ -0,0 +1,47 @@
from pathlib import Path
from nonebot import require
require("nonebot_plugin_localstore")
import json
from nonebot_plugin_localstore import get_data_file
memory_path = get_data_file("marshoai", "memory.json")
if not Path(memory_path).exists():
with open(memory_path, "w", encoding="utf-8") as f:
json.dump({}, f, ensure_ascii=False, indent=4)
print(memory_path)
async def write_memory(memory: str, user_id: str):
with open(memory_path, "r", encoding="utf-8") as f:
memory_data = json.load(f)
memorys = memory_data.get(user_id, [])
memorys.append(memory)
memory_data[user_id] = memorys
with open(memory_path, "w", encoding="utf-8") as f:
json.dump(memory_data, f, ensure_ascii=False, indent=4)
return "记忆已经保存啦~"
async def read_memory(user_id: str):
with open(memory_path, "r", encoding="utf-8") as f:
memory_data = json.load(f)
memorys = memory_data.get(user_id, [])
if not memorys:
return "好像对ta还没有任何记忆呢~"
return "这些是有关ta的记忆" + "\n".join(memorys)
async def organize_memories():
with open(memory_path, "r", encoding="utf-8") as f:
memory_data = json.load(f)
for i in memory_data:
...
# TODO 用大模型对记忆进行整理

View File

@@ -0,0 +1,46 @@
[
{
"type": "function",
"function": {
"name": "marshoai_memory__write_memory",
"description": "如果在上下中你看见并觉得应该记住的人的行为与事件请调用这个函数并将记忆内容写入。请尽量每次都调用总结ta的习惯、爱好和性格,以及你对ta的印象和ta对你的印象",
"parameters": {
"type": "object",
"properties": {
"memory": {
"type": "string",
"description": "你想记住的内容,概括并保留关键内容。"
},
"user_id": {
"type": "string",
"description": "你想记住的人的id。"
}
}
},
"required": [
"memory",
"user_id"
]
}
},
{
"type": "function",
"function": {
"name": "marshoai_memory__read_memory",
"description": "每当你想要获取更多有关某人的信息的时候,请调用这个函数。",
"parameters": {
"type": "object",
"properties": {
"user_id": {
"type": "string",
"description": "你想获取的人的id。"
}
}
},
"required": [
"user_id"
]
}
}
]

View File

@@ -1,2 +0,0 @@
async def write_memory(memory: str):
return ""

View File

@@ -1,21 +0,0 @@
[
{
"type": "function",
"function": {
"name": "marshoai_memory__write_memory",
"description": "当你想记住有关与你对话的人的一些信息的时候,调用此函数。",
"parameters": {
"type": "object",
"properties": {
"memory": {
"type": "string",
"description": "你想记住的内容,概括并保留关键内容。"
}
}
},
"required": [
"memory"
]
}
}
]

View File

@@ -16,6 +16,8 @@ from nonebot.log import logger
from nonebot_plugin_alconna import Image as ImageMsg from nonebot_plugin_alconna import Image as ImageMsg
from nonebot_plugin_alconna import Text as TextMsg from nonebot_plugin_alconna import Text as TextMsg
from nonebot_plugin_alconna import UniMessage from nonebot_plugin_alconna import UniMessage
from openai import AsyncOpenAI
from zhDateTime import DateTime
from .config import config from .config import config
from .constants import * from .constants import *
@@ -27,7 +29,7 @@ loaded_target_list = [] # 记录已恢复备份的上下文的列表
# noinspection LongLine # noinspection LongLine
chromium_headers = { chromium_headers = {
"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36" "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:134.0) Gecko/20100101 Firefox/134.0"
} }
@@ -90,7 +92,8 @@ async def make_chat(
参数: 参数:
client: 用于与AI模型进行通信 client: 用于与AI模型进行通信
msg: 消息内容 msg: 消息内容
model_name: 指定AI模型名""" model_name: 指定AI模型名
tools: 工具列表"""
return await client.complete( return await client.complete(
messages=msg, messages=msg,
model=model_name, model=model_name,
@@ -101,6 +104,29 @@ async def make_chat(
) )
async def make_chat_openai(
client: AsyncOpenAI,
msg: list,
model_name: str,
tools: Optional[list] = None,
):
"""使用 Openai SDK 调用ai获取回复
参数:
client: 用于与AI模型进行通信
msg: 消息内容
model_name: 指定AI模型名
tools: 工具列表"""
return await client.chat.completions.create(
messages=msg,
model=model_name,
tools=tools,
temperature=config.marshoai_temperature,
max_tokens=config.marshoai_max_tokens,
top_p=config.marshoai_top_p,
)
def get_praises(): def get_praises():
global praises_json global praises_json
if praises_json is None: if praises_json is None:
@@ -220,6 +246,13 @@ def get_prompt():
if config.marshoai_enable_praises: if config.marshoai_enable_praises:
praises_prompt = build_praises() praises_prompt = build_praises()
prompts += praises_prompt prompts += praises_prompt
if config.marshoai_enable_time_prompt:
current_time = DateTime.now().strftime("%Y.%m.%d %H:%M:%S")
current_lunar_date = (
DateTime.now().to_lunar().date_hanzify()[5:]
) # 库更新之前使用切片
time_prompt = f"现在的时间是{current_time},农历{current_lunar_date}"
prompts += time_prompt
marsho_prompt = config.marshoai_prompt marsho_prompt = config.marshoai_prompt
spell = SystemMessage(content=marsho_prompt + prompts).as_dict() spell = SystemMessage(content=marsho_prompt + prompts).as_dict()
return spell return spell

View File

@@ -1,15 +1,10 @@
{ {
"type": "module", "type": "module",
"devDependencies": { "devDependencies": {"vitepress": "^1.5.0", "vitepress-sidebar": "^1.30.2"},
"vitepress": "^1.5.0", "scripts": {
"vitepress-sidebar": "^1.30.2" "docs:dev": "vitepress dev docs --host",
}, "docs:build": "vitepress build docs",
"scripts": { "docs:preview": "vitepress preview docs"
"docs:dev": "vitepress dev docs --host", },
"docs:build": "vitepress build docs", "dependencies": {"vue": "^3.5.13"}
"docs:preview": "vitepress preview docs" }
},
"dependencies": {
"vue": "^3.5.13"
}
}

View File

@@ -9,7 +9,7 @@ authors = [
{ name="LiteyukiStudio", email = "support@liteyuki.icu"} { name="LiteyukiStudio", email = "support@liteyuki.icu"}
] ]
dependencies = [ dependencies = [
"nonebot2>=2.2.0", "nonebot2>=2.4.0",
"nonebot-plugin-alconna>=0.48.0", "nonebot-plugin-alconna>=0.48.0",
"nonebot-plugin-localstore>=0.7.1", "nonebot-plugin-localstore>=0.7.1",
"zhDatetime>=1.1.1", "zhDatetime>=1.1.1",
@@ -26,7 +26,9 @@ dependencies = [
"aiofiles>=24.1.0", "aiofiles>=24.1.0",
"sumy>=0.11.0", "sumy>=0.11.0",
"azure-ai-inference>=1.0.0b6", "azure-ai-inference>=1.0.0b6",
"watchdog>=6.0.0" "watchdog>=6.0.0",
"nonebot-plugin-apscheduler>=0.5.0",
"openai>=1.58.1"
] ]
license = { text = "MIT, Mulan PSL v2" } license = { text = "MIT, Mulan PSL v2" }
@@ -76,3 +78,6 @@ dev = [
test = [ test = [
"nonebug>=0.4.3", "nonebug>=0.4.3",
] ]
[tool.ruff.lint]
ignore = ["E402"]