Security audit
Security checks for vulnerabilities and agentic risk
This is a coherent fiction-writing assistant skill with no executable code, remote payloads, or hidden data handling; its main caveats are broad activation wording and Chinese-first guidance.
Install this if you want a Chinese-first workflow for novel or serialized-fiction drafting. Be aware it may create or update local story files such as story-bible.md, and its broad triggers may make it activate for some writing or polishing requests unless your tool lets you scope skill use.
Referenced artifact was not completely inspected
- **Cursor**：Cursor 用的是 `.cursor/rules/*.mdc` 而不是 `SKILL.md`。把本文件正文（frontmatter 之后的部分）复制进 `.cursor/rules/novel-writing.mdc`，文件头换成：
The trigger definition is overly broad because it includes many generic writing-related requests plus the open-ended phrase 'or similar phrases,' which can cause the skill to activate in ambiguous contexts. Over-broad activation can route unrelated prompts into this skill, override more appropriate skills or baseline handling, and increase the chance of prompt-scope confusion or unintended instruction application.
The usage section expands activation through broad keyword matching across Chinese and English without defining exclusions, making accidental invocation more likely. In a skill-routing system, this can misclassify ordinary writing or review requests as fiction-workflow tasks, causing inappropriate file creation, context loading, or structured guidance to be applied where it does not belong.
This markdown file is written entirely in Chinese and presents the checklist as the skill content without indicating that users may choose another language. Under the policy rule, forcing a specific language without user opt-in can be a natural-language policy violation.
The checklist is written as the default operating procedure in Chinese and references a Chinese-only output format, which can steer the agent to respond in Chinese without checking the user's language preference. This creates a user-choice and policy-compliance issue: the skill may override the user's requested language or reduce accessibility for non-Chinese users.
This reference file is written entirely in Chinese and instructs the reviewer in Chinese without offering a language-selection mechanism or documenting that the skill is intentionally limited to Chinese-speaking users. In a broadly triggered writing skill, this can cause the agent to respond in an unexpected language, reducing usability, accessibility, and potentially causing the user to miss important review guidance.
No suspicious patterns detected.