Security audit
Security checks for vulnerabilities and agentic risk
This skill is a straightforward image OCR helper that processes user-provided images and returns extracted text without hidden execution or persistence.
Before installing, note that this skill will inspect the contents of images you provide and may create a Markdown file if requested. Avoid uploading sensitive images unless you are comfortable having their text extracted in the current session.
The skill description is broad enough to overlap with many generic image-processing or OCR requests, which can cause unintended activation. While this is not a direct code-execution issue, misrouting user requests to the wrong skill can lead to incorrect handling of user data or unexpected tool use.
The trigger section defines when the skill should activate but does not define clear boundaries for when it should not. This ambiguity increases the chance of accidental invocation on adjacent workflows, especially broad 'analyze these images' requests, which can degrade security predictability and cause unnecessary access to image contents.
The title and the entire specification require output in Chinese, including fixed headings such as "图片文字提取结果". This imposes a specific language/locale on users without opt-in or any stated regional/business justification, which matches the language-policy violation criteria.
No suspicious patterns detected.
