baoyu-post-to-wechat
Pass
Audited by Gen Agent Trust Hub on Aug 14, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONCREDENTIALS_UNSAFE
Full Analysis
- [REMOTE_CODE_EXECUTION]: Automated scans detected a shell command for installing the Bun runtime (curl -fsSL https://bun.sh/install | bash). This is the official installation method for the runtime required by the skill and originates from a well-known service.
- [EXTERNAL_DOWNLOADS]: The skill fetches remote images referenced in markdown files to prepare them for uploading to WeChat. It also loads syntax highlighting and diagram rendering libraries from official project CDNs.
- [COMMAND_EXECUTION]: The skill automates UI tasks by executing local TypeScript and Swift scripts. It uses the Chrome DevTools Protocol (CDP) to interface with the WeChat web editor and utilizes system automation tools like osascript or xdotool for clipboard and keystroke simulation.
- [CREDENTIALS_UNSAFE]: The skill requires WeChat API credentials for posting. It follows secure development practices by instructing users to manage these keys through local environment variables or configuration files rather than hardcoding them.
- [DYNAMIC_EXECUTION]: The skill generates and executes temporary Swift scripts on macOS to handle specialized clipboard operations. It also dynamically loads the juice and infographic libraries during the markdown rendering process.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied markdown and HTML files. While it uses standard parsers, the ingestion of untrusted content combined with automated browser interactions based on that content represents a vulnerability surface.
Audit Metadata