Security audit
Security checks for vulnerabilities and agentic risk
This is a Chinese-language prompt-writing skill for AI video storyboards and does not install code or request sensitive access.
Use this skill if you want Chinese-formatted AI-video storyboard prompts. Be mindful that it may recommend uploading product or character reference assets to external video/model platforms, so only provide assets you are comfortable sharing there.
The README is entirely written as a Chinese-language skill specification and presents the output format and behavior as fixed, with no indication that users can choose another language or that the skill is limited to a Chinese-only region or audience. Under the stated policy, forcing a specific language without user opt-in is a natural-language policy violation.
The skill description and behavior are written entirely in Chinese and do not disclose any locale limitation or provide a language fallback. In a multi-language agent environment, this can cause user confusion, incorrect prompting, or unsafe misunderstandings if a user expects another language and the model silently forces Chinese output.
The mandated output section requires Chinese-only field labels and phrasing with no locale negotiation. This rigid formatting can mislead downstream users or systems that assume outputs will match user language, increasing the chance of operational mistakes, mistranslation, or failed parsing in mixed-language workflows.
The fixed V2 template hardcodes Chinese-only template text without documenting locale constraints. While not a code-execution risk, it is a genuine security-quality issue because it can produce misunderstood instructions or unusable outputs in cross-lingual settings, especially where precise prompt semantics matter.
This markdown file is entirely written in Chinese and presents the scoring rubric as the required format, with no indication that users may choose another language or that the rubric is limited to a Chinese-specific context. Under the language/locale policy, forcing a specific language without opt-in is a natural-language policy violation.
The specification repeatedly directs users to upload the file to DeepSeek, 豆包, or other models and is written entirely in Chinese, implying a fixed language/locale workflow. There is no explicit user opt-in, alternative locale handling, or justification that this skill is intended only for a Chinese-language context.
This markdown test-case file presents all instructions, inputs, and expected checks exclusively in Chinese. Under the policy-violation category, forcing a specific language without user opt-in can be a natural-language policy issue when no alternative language or explicit locale constraint is documented.
