Security audit
Security checks for vulnerabilities and agentic risk
This is a straightforward image watermarking skill with disclosed local file processing and no hidden network, credential, persistence, or privilege behavior.
Install only if you are comfortable with a Chinese-language watermarking tool that processes local image files you explicitly provide. Review output paths before batch use, because it recursively scans the chosen folder and writes generated images to the chosen output directory; batch mode may need a bug fix before it works as documented.
The file name, description, headings, usage guidance, and labels are written in Chinese, which effectively forces a specific language for users. Under the policy, language constraints should either provide user opt-in/choice or be explicitly justified as region-specific.
The script’s title, docstrings, usage output, parameter conventions, and error messages are all in Chinese, and the accepted position values are also Chinese-only. This imposes a specific language/locale on users without opt-in or documented justification, which matches the natural-language policy violation criteria.
The module docstring presents the tool entirely in Chinese, and the rest of the CLI help and status messages are likewise Chinese-only. For a general-purpose skill, this imposes a specific language on users without opt-in or justification that the skill is intended only for a Chinese-speaking or region-specific context.
The file presents itself as an image watermark tool with batch support, and the batch subcommand builds a result object at L185. However, the final exit path at L188 unconditionally reads result["status"], which is only set for text/image modes, so batch mode contradicts the documented '批量处理' behavior by raising a KeyError instead of completing normally.
No suspicious patterns detected.
