Wechat Workflow

微信公众号虾发布工作流 — AI 生成封面图 + 内容创作（去AI味+爆款结构）+ 一键发布草稿箱 + 搜狗验证收录 + 数据追踪 + 选题建议。全链路自动化。

This skill appears to do what it says: publish Markdown to a WeChat draft, check Sogou indexing, and track articles. Before installing/use: 1) Understand it will store your WECHAT_APP_ID and WECHAT_APP_SECRET in ~/.openclaw/workspace/secrets.json (file saved with 0600 permissions) and will also parse TOOLS.md as a fallback — only proceed if you are comfortable storing these creds locally. 2) The publish script may auto-run `npm install -g @wenyan-md/cli` if the 'wenyan' CLI is missing; that performs a global npm install (network access and write permissions to global npm dirs). If you prefer, install @wenyan-md/cli manually in a controlled environment or run the scripts in a container/VM. 3) Review the behavior of the wenyan CLI and the khazix-writer repo (the skill suggests installing khazix-writer) before granting credentials. 4) Note the small documentation mismatch: SKILL metadata lists only 'wenyan' as a required binary, but the scripts require Node/npm; ensure Node/npm are present or handle the install manually. 5) If you have sensitive production WeChat credentials, consider creating a dedicated, limited-scope account for automation or audit the code and test in an isolated environment first.

Type: OpenClaw Skill Name: wechat-workflow Version: 1.3.0 The skill bundle automates WeChat article publishing and tracking but contains high-risk behaviors and potential vulnerabilities. Specifically, 'publish_and_verify.py' automatically performs a global system modification by running 'npm install -g @wenyan-md/cli' if the dependency is missing, and it utilizes 'shell=True' for command execution on Windows, which is a known shell injection risk. While these actions are documented and support the stated purpose, the combination of system-level changes and risky execution patterns warrants a suspicious classification. Credential management is handled locally in 'secrets.json' with appropriate file permissions, and external network calls to 'api.weixin.qq.com', 'weixin.sogou.com', and 'ifconfig.me' appear legitimate for the workflow.