← Back to Skills Marketplace
163
Downloads
1
Stars
0
Active Installs
11
Versions
Install in OpenClaw
/install wechatwindowllmtranslate
Description
微信聊天窗口自动监控、翻译和智能回复工具，集成 Qwen 大语言模型和百度翻译 API。
Usage Guidance
这是一个功能与要求基本一致的 Windows 自动化脚本——在决定安装/运行前请注意：
- 隐私与数据外发：脚本会读取微信消息并把文本发送到第三方服务（Qwen 模型与百度翻译），因此任何敏感或个人信息可能被上报。仅在你接受这些外发行为时使用。
- API Key 管理：为 Qwen/百度翻译 使用最小权限或测试账户；不要使用主账号或高权限密钥。避免将密钥提交到公共仓库。
- 检查并确认 API 域名：脚本提供了默认 QWEN_BASE_URL，请替换为你信任的 API 端点并验证其合法性。
- 运行环境建议：先在受控/沙箱环境中测试（使用虚假或受限的密钥），确认坐标和行为不会误操作其他窗口或发送错误信息。
- 日志与网络监控：如果担心外发内容，可在首次运行时监控网络请求或通过本地防火墙限制对不信任域的出站访问。
- 代码审查：如有顾虑，审阅 wechatai.py 的剩余部分（截断处）以确认没有隐蔽的远程上载、命令执行或敏感文件读取逻辑。
如果你不接受脚本将聊天内容上传给外部服务或不能为密钥设置受限权限，请不要安装/运行该技能。
Capability Analysis
Type: OpenClaw Skill
Name: wechatwindowllmtranslate
Version: 1.1.0
The skill provides automated WeChat monitoring and AI-powered replies using high-risk capabilities such as programmatic UI control (`pyautogui`), clipboard access (`pyperclip`), and window manipulation (`win32gui`). The script (`wechatai.py`) identifies the WeChat window, scrapes chat content via clipboard copying, and automatically sends generated responses by simulating keystrokes. While these actions are aligned with the stated purpose of a chat automation tool, the broad access to the clipboard and the ability to control user input on a Windows system represent a significant security risk. No evidence of intentional malice or unauthorized data exfiltration was found, but the tool's design requires handling sensitive API keys and chat data in a high-privilege manner.
Capability Assessment
Purpose & Capability
技能名/描述（监控微信窗口、翻译并调用 Qwen 生成回复）与代码和 SKILL.md 中的行为一致：脚本通过 pyautogui/pyperclip 操作窗口与剪贴板，调用 Qwen 与百度翻译 API。所需 env（QWEN_*、BAIDU_*）与用途匹配。
Instruction Scope
SKILL.md 明确指示启动 wechatai.py，轮询微信窗口、双击复制消息然后发送回复；因此脚本将访问并操作屏幕、剪贴板和窗口，并把消息内容上传到外部 LLM/翻译服务——这是功能所需但带来隐私/数据外发风险。SKILL.md 也建议把 API 密钥放到 openclaw.json 或环境变量，且包含运行与坐标调整说明。
Install Mechanism
这是 instruction-only 的技能（无 install spec），仓内有 Python 脚本和 requirements.txt；依赖为常见库（requests、pyautogui、pyperclip、pywin32），无不明下载 URL 或可执行安装步骤，安装风险较低。
Credentials
需要的环境变量（QWEN_BASE_URL/KEY/URL_PATH 与 BAIDU_APPID/APPKEY）与技能功能直接相关。注意 SKILL.md/package.json 还引用 BAIDU_ENDPOINT/BAIDU_PATH 和可选 mbti_selected 等配置；这些在 registry metadata 中作为非必需/示例出现，整体合理，但请确认 QWEN_BASE_URL 指向可信域（脚本默认值为 api-ai.gitcode.com，应按需替换）。
Persistence & Privilege
flags 表示 always:false，技能不会被强制常驻。技能允许模型自主调用（平台默认），但没有请求修改其它技能或系统范围配置，权限要求仅限于屏幕控制和剪贴板访问（脚本运行时会执行这些操作）。
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install wechatwindowllmtranslate - After installation, invoke the skill by name or use
/wechatwindowllmtranslate - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.1.0
- 更新 SKILL.md 中的坐标调整说明，去除了具体坐标示例，改为通用描述，适应更多用户环境
- 增加了快捷退出机制
- 版本号从 1.0.3 升级为 1.0.10 并同步到文档底部
- 其余使用说明、依赖、配置及功能未改动
v1.0.10
- Clarified usage instructions to include steps for logging into desktop WeChat and how to exit by moving the mouse to the far left.
- Updated "使用方法" section, adding practical start-up and exit instructions for different run methods.
- No changes to core functionality, dependencies, or configuration requirements.
- Documentation improvements for easier setup and operation.
v1.0.9
Version 1.0.9
- No file or documentation changes detected in this release.
v1.0.8
- Added instructions for configuring environment variables in openclaw.json, including options for mbti_selected and refresh interval.
- Clarified steps for exiting by closing the WeChat window.
- No changes to skill features or primary environment variable use.
- Documentation updates only; core logic unchanged.
v1.0.7
- Added a feature highlight for optional MBTI 16 personality chat styles (“个性化聊天”) in the feature list.
- No file changes or code updates detected; documentation only.
v1.0.6
- No visible file changes detected in this version.
- No updates or modifications found in SKILL.md or source files.
v1.0.5
- No file changes detected in this release.
- Documentation and metadata remain unchanged from previous versions.
- Functionality is identical to previous version 1.0.0.
v1.0.4
- 增加了对 MBTI 个性化回复的可选支持，优化了个性化回复体验
- 工作原理说明中新增使用 Qwen 模型生成 MBTI 个性回答的流程
- 聊天窗口滚动起始坐标由 (470, 650) 调整为 (430, 650)
- 文档描述优化，增加对 MBTI 个性化功能的说明
v1.0.3
- Renamed main module file from weichatai.py to wechatai.py.
- No user-facing functional changes.
v1.0.2
- Renamed implementation file from wechatai.py to weichatai.py.
- No changes to documentation or Skill metadata.
- No functional or user-facing updates beyond the file rename.
v1.0.0
桌面微信窗口最大化，否则根据实际情况调整屏幕坐标
Metadata
Frequently Asked Questions
What is 微信智能聊天(MBTI版)?
微信聊天窗口自动监控、翻译和智能回复工具，集成 Qwen 大语言模型和百度翻译 API。 It is an AI Agent Skill for Claude Code / OpenClaw, with 163 downloads so far.
How do I install 微信智能聊天(MBTI版)?
Run "/install wechatwindowllmtranslate" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is 微信智能聊天(MBTI版) free?
Yes, 微信智能聊天(MBTI版) is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does 微信智能聊天(MBTI版) support?
微信智能聊天(MBTI版) is cross-platform and runs anywhere OpenClaw / Claude Code is available (win32).
Who created 微信智能聊天(MBTI版)?
It is built and maintained by chaoyi001 (@chaoyi001); the current version is v1.1.0.
More Skills