Security audit
Security checks for vulnerabilities and agentic risk
This is a coherent Chinese novel-writing helper that creates outline and chapter files, with no evidence of hidden execution or data theft.
Before installing, expect the skill to guide a Chinese web-novel workflow and create one markdown file per chapter under ./novels/小说标题/. Choose or monitor the output folder if storage, filenames, or overwriting existing drafts matter to you.
The trigger phrases are very broad generic writing requests, so the skill may activate for ordinary brainstorming or editing asks where the user did not intend a multi-step novel-generation workflow. This can cause unintended data collection, workflow hijacking, and downstream file-generation behavior that exceeds user expectations.
The trigger section does not define clear activation boundaries, disambiguation logic, or when the skill should refuse to engage, making accidental invocation more likely. In this context, accidental activation is more dangerous because the skill can transition into automated content generation and persistence to disk.
The skill instructs automatic chapter-by-chapter saving to the filesystem without a clear upfront warning, explicit user consent for writes, or constraints on write scope. Even though the content domain is benign, silent or assumed file creation can surprise users, consume storage, overwrite existing content, or create unwanted persistent artifacts.
The document title and all instructional content are written entirely in Chinese, with no indication that the skill supports other languages or that Chinese is an intentional, documented locale restriction. Under the policy, forcing a specific language without user opt-in can be a natural-language policy violation.
No suspicious patterns detected.
