← 返回 Skills 市场
dragon015

Wechat Mcp

作者 dragon015 · GitHub ↗ · v1.0.0
cross-platform ⚠ suspicious
2669
总下载
4
收藏
13
当前安装
1
版本数
在 OpenClaw 中安装
/install wechat-mcp
功能描述
Windows 电脑端微信消息监控与发送 MCP,实现在微信上给指定联系人发送消息
安全使用建议
This skill automates your Windows desktop to send WeChat messages by simulating clicks and clipboard operations. Before installing or running it: - Review the code yourself (or ask the author) because SKILL.md claims a contact-search/send_message_to_contact flow that is not implemented; expect only send_message_to_current behavior. - Be aware the code disables pyautogui's FAILSAFE (pyautogui.FAILSAFE = False). Consider changing that to True to allow aborting automation by moving the mouse to a screen corner. - The skill writes screenshot files (verify.png, input_verify.png, result.png) into the skill directory — ensure that is acceptable for your environment. - requirements.txt lists pytesseract (and Tesseract requires a separate system binary). Confirm whether OCR is needed and install/enable it only if required. - Because this automates the GUI and clipboard, test in a safe environment (with a test WeChat account) to avoid sending messages to the wrong contact. If you need higher assurance, ask the publisher to: (1) fix the documentation to match implemented functions, (2) remove or justify unnecessary dependencies, (3) re-enable pyautogui failsafe or add an explicit emergency-stop mechanism, and (4) declare any files the skill will create.
功能分析
Type: OpenClaw Skill Name: wechat-mcp Version: 1.0.0 The skill is classified as suspicious due to the inherent risks associated with GUI automation and the explicit disabling of `pyautogui.FAILSAFE` in `server.py`. While the stated purpose of automating WeChat message sending appears legitimate, the use of `pyautogui` grants broad control over the user's desktop, including simulating mouse clicks and keyboard presses, which could lead to unintended actions if the environment changes or the script's assumptions are incorrect. The disabling of the failsafe mechanism removes a critical safety feature, increasing the risk of a runaway script. Additionally, the script saves screenshots (`verify.png`, `input_verify.png`, `result.png`) to the local directory, which could inadvertently capture and store sensitive information from the screen.
能力评估
Purpose & Capability
The package purpose (monitoring and sending WeChat messages on Windows) aligns with the code: pyautogui/pygetwindow/ImageGrab-based desktop automation. However SKILL.md and examples claim functions/features that are not present (e.g., send_message_to_contact and contact-search/open flow). requirements.txt also lists pytesseract even though server.py does not use it. These mismatches suggest the docs and code are out of sync.
Instruction Scope
Runtime instructions tell the agent to install Python GUI automation libs and call functions like send_message_to_contact, but the actual runtime entry points in server.py are send_message_to_current and get_wechat_status (no send_message_to_contact). The code performs GUI clicks, clipboard operations, and saves screenshots (verify.png, input_verify.png, result.png) to the skill directory — the SKILL.md does not clearly document the file writes. The skill does not access external networks or credentials, but the discrepancy between documented and implemented behavior is a scope/integrity problem.
Install Mechanism
There is no install spec (instruction-only), which is low risk. requirements.txt lists typical GUI automation packages. It also lists pytesseract (and the SKILL.md does not instruct installing Tesseract OCR engine), which may be unnecessary or missing a required external binary (tesseract.exe) on Windows — the README and instructions should clarify this dependency.
Credentials
The skill requests no environment variables or credentials and does not call external services. It does interact with the system GUI, clipboard (pyperclip), and filesystem (writes PNG files to the skill directory) which is proportionate to desktop automation but worth noting because these actions can leak or alter local data if misdirected.
Persistence & Privilege
The skill is not marked always:true and does not modify other skills or system-wide configs. It runs an stdin JSON-RPC loop when executed, which can autonomously perform GUI actions if invoked — this is expected for an MCP-style tool but increases impact if used without oversight.
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-mcp
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /wechat-mcp 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
v1.0.0
WeChat MCP v1.0.0 初始发布 - 实现 Windows 端微信消息监控与发送能力 - 支持截图、搜索并打开联系人、发送消息功能 - 可通过 MCP 协议或 Python 调用 - 支持独立聊天窗口识别与消息发送 - 包含微信窗口状态获取工具
元数据
Slug wechat-mcp
版本 1.0.0
许可证
累计安装 13
当前安装数 13
历史版本数 1
常见问题

Wechat Mcp 是什么?

Windows 电脑端微信消息监控与发送 MCP,实现在微信上给指定联系人发送消息. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 2669 次。

如何安装 Wechat Mcp?

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

Wechat Mcp 是免费的吗?

是的,Wechat Mcp 完全免费(开源免费),可自由下载、安装和使用。

Wechat Mcp 支持哪些平台?

Wechat Mcp 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 Mcp?

由 dragon015(@dragon015)开发并维护,当前版本 v1.0.0。

💬 留言讨论