← 返回 Skills 市场
chenjunlei1989-glitch

WeChat Auto Reply

作者 chenjunlei1989-glitch · GitHub ↗ · v1.0.1 · MIT-0
cross-platform ⚠ suspicious
98
总下载
1
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install wechat-auto-reply-mac
功能描述
微信消息自动发送/半自动回复。主动发送时,按“搜索联系人→单聊直接 Enter 进入聊天;群聊先识别群聊分组再定位目标项→粘贴消息→发送”的逻辑执行。适用于 macOS + 微信桌面版环境,需本机完成权限和依赖配置。使用方式:wechat-auto-reply "联系人名称" 或 wechat-auto-repl...
安全使用建议
This package appears to implement WeChat UI automation on macOS, which reasonably needs Accessibility permission, cliclick, screencapture, and Vision OCR. Before installing: - Do NOT grant Accessibility permission until you trust the scripts. UI automation can read and control your screen and keyboard. - The package references a wechat-dm.applescript file but that file is not included—installation as provided will fail. Ask the author or the tap for the missing AppleScript and inspect it before running. - The README suggests installing from a third-party Homebrew tap (bjdzliu/openclaw). Only add and install from taps you trust and inspect the formula contents. - Inspect the AppleScript and any Homebrew formula for network or credential usage. The included Python and shell files do not perform network calls or credential exfiltration, but the missing AppleScript could change that behavior. - Be prepared to calibrate hard-coded click coordinates and to install cliclick and pyobjc dependencies in a safe environment first. The repo also contains developer-local paths (e.g., /Users/chuck/.venvs/...), a sign of incomplete packaging. If you want to proceed, request the missing wechat-dm.applescript and the Homebrew formula source, review them, and test in a non-critical account or VM before granting Accessibility to the scripts.
功能分析
Type: OpenClaw Skill Name: wechat-auto-reply-mac Version: 1.0.1 The skill automates WeChat on macOS using high-risk techniques including screen capture (`screencapture`), OCR (Vision Framework), and UI manipulation via `cliclick` and AppleScript. While these actions align with the stated 'auto-reply' purpose, the Python scripts (`state_machine_final.py` and `state_machine_test.py`) contain vulnerabilities where user-controlled strings (contact names/messages) are interpolated directly into AppleScript commands without sanitization, creating a risk of local injection. The inclusion of hardcoded developer-specific file paths (e.g., `/Users/chuck/`) and fixed screen coordinates suggests the code is unvetted and may behave unpredictably in different environments.
能力评估
Purpose & Capability
The name/description (WeChat auto reply on macOS) aligns with the included files and runtime actions: AppleScript + osascript calls, macOS Vision OCR, screencapture, and cliclick-driven UI automation. Requested permissions (Accessibility) and tools (cliclick, Vision, screencapture) are appropriate for the stated goal.
Instruction Scope
SKILL.md and the shell wrapper expect an AppleScript file wechat-dm.applescript to exist and be used for the main CLI; that file is referenced in multiple places but is not present in the package. Some scripts also embed developer-local paths (e.g., VENV_PY = "/Users/chuck/.venvs/pyobjc/bin/python") and hard-coded click coordinates which indicate per-machine calibration is required. These inconsistencies mean the provided instructions may not work as-is and could lead users to run or fetch additional files from external sources.
Install Mechanism
There is no formal install spec in the registry package (instruction-only install), but SKILL.md recommends a third-party Homebrew tap (bjdzliu/openclaw). That recommendation is plausible but not enforced by the package metadata—users would need to trust and inspect that tap. No network endpoints or downloads are present in the included code files, lowering code-execution risk from remote fetches, but the brew tap advice introduces a potential external trust decision.
Credentials
The skill requests no environment variables or credentials, which is appropriate. Caveats: scripts expect cliclick at /usr/local/bin and use /usr/sbin/screencapture and macOS Vision—the absolute paths and a leftover developer venv path (/Users/chuck/...) are packaging artifacts that should be adjusted for the target machine. The skill requires Accessibility permission, which is necessary for UI automation but is a high-privilege capability that users must consciously grant.
Persistence & Privilege
The skill does not declare always:true and does not request to modify other skills or global agent settings. It operates locally and requires user-granted macOS Accessibility permission; it does not request background persistence beyond normal installation locations described in SKILL.md.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install wechat-auto-reply-mac
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /wechat-auto-reply-mac 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.1
基于wechat-auto-reply优化的版本
v1.0.0
基于wechat-auto-reply优化的版本
元数据
Slug wechat-auto-reply-mac
版本 1.0.1
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 2
常见问题

WeChat Auto Reply 是什么?

微信消息自动发送/半自动回复。主动发送时,按“搜索联系人→单聊直接 Enter 进入聊天;群聊先识别群聊分组再定位目标项→粘贴消息→发送”的逻辑执行。适用于 macOS + 微信桌面版环境,需本机完成权限和依赖配置。使用方式:wechat-auto-reply "联系人名称" 或 wechat-auto-repl... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 98 次。

如何安装 WeChat Auto Reply?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install wechat-auto-reply-mac」即可一键安装,无需额外配置。

WeChat Auto Reply 是免费的吗?

是的,WeChat Auto Reply 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

WeChat Auto Reply 支持哪些平台?

WeChat Auto Reply 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 WeChat Auto Reply?

由 chenjunlei1989-glitch(@chenjunlei1989-glitch)开发并维护,当前版本 v1.0.1。

💬 留言讨论