通过微信控制openclaw

在 OpenClaw 中提供 WeChat 回调接入、群私聊会话路由、消息发送与图片识别入口能力。

This package appears to do what it says: run a local FastAPI webhook that forwards WeChat messages to OpenClaw via the CLI. Before installing, consider: - Privacy: by default the code uses api.wechatapi.net as the WeChat API backend. Messages and images may be proxied through that third-party service. If you have sensitive data, either host your own WeChat API backend or confirm the third party's privacy/security policy. - Credentials: WX_API_TOKEN is required — treat it like any service token. Do not paste a privileged token into unknown code or public repos. - Public exposure: PUBLIC_URL must be reachable from the internet for callbacks. Exposing a public callback can accept inbound traffic; run behind TLS/reverse-proxy or on an isolated host if needed. - CLI execution: the gateway invokes the OpenClaw CLI per message (subprocess). Ensure OPENCLAW_BIN points to the intended binary and run in an environment where executing that binary is safe. An attacker replacing that binary could cause arbitrary code execution. - Storage: the skill writes config.ini, logs, and saved images locally. Monitor disk usage and clear stored images if they contain sensitive content. - Review & sandboxing: if you are not 100% comfortable, review the full main.py (it’s included) or run it in an isolated VM/container before production use. Verify/pin dependency versions when installing the required Python packages. If you want a higher-confidence assessment, provide the full (non-truncated) main.py content so I can scan for any hidden network endpoints, unusual subprocess calls, or data-exfiltration patterns.

Type: OpenClaw Skill Name: wechat-gateway Version: 1.0.0 The skill bundle provides a WeChat gateway for OpenClaw but contains a path traversal vulnerability in `main.py` within the `save_incoming_image_from_base64` function, where an unsanitized `msg_id` from an external API callback is used to construct local file paths. It also implements a weak 'magic phrase' authentication mechanism ('我是你的主人') that allows any user to add themselves to the administrative whitelist. While the code performs high-risk actions such as executing shell commands via `subprocess.run` and communicating with a third-party API (wechatapi.net), these appear to be functional requirements rather than intentional malice.