Security audit
Security checks for vulnerabilities and agentic risk
This is a Markdown guide for keeping AI-generated characters visually consistent, with no executable code or hidden automation.
Install only if you want a Chinese-language character-consistency workflow. If using real performers or models, keep reference photos and meta.json private, get consent, and delete or minimize sensitive appearance details when no longer needed.
The manifest description and title are presented in Chinese, framing the skill for Chinese-language use without any indication that users may choose another language or locale. This is a natural-language policy concern because the file does not offer opt-in language selection or document a justified region-specific restriction.
The guide explicitly recommends storing personally identifying and sensitive attributes in meta.json, including age appearance, gender, skin tone, body build, distinguishing marks, and wardrobe details, but provides no privacy, consent, retention, or access-control guidance. In a character-consistency workflow this data may often describe real performers or models, so centralizing it in a structured file increases the risk of privacy leakage, profiling, or misuse if the project files are shared, synced, or exposed.
The title and introduction present the material in a mixed Chinese-English format, and the document does not indicate that users can choose another language or locale. Under the policy, forcing a specific language or locale without opt-in can be a natural-language policy concern.
No suspicious patterns detected.
