Security audit
Security checks for vulnerabilities and agentic risk
This skill is a local Chinese novel chapter quality checker that reads user-supplied chapter files and reports quality metrics without persistence, network access, or destructive behavior.
Install this only if you want a Chinese novel chapter checker. When using directory mode, point it only at the chapter folder you intend to audit, because the script will read matching markdown files in that directory for analysis.
The skill instructs users to run a local Python script against arbitrary files and directories, but the manifest does not declare any explicit tool scope or file access boundaries. In an agent environment, missing permission declarations can lead to overly broad file-read behavior or unclear authorization expectations, increasing the risk of unintended access to user files during chapter scans.
The trigger phrases include generic expressions such as '审核一下' and '看看这章行不行', which are broad enough to match ordinary conversation and may activate the skill when the user did not intend a quality audit. Unintended activation can expose local chapter files to analysis routines or cause the agent to apply restrictive auditing behavior in the wrong context.
The documented quality criteria require ‘中文纯度（>95%汉字）’, which effectively forces Chinese-language output as a normative requirement. The file does not indicate that this is optional, user-selected, or limited to a clearly documented region-specific workflow.
This is a natural-language locale policy issue because the file presents the skill exclusively in Chinese for its title, usage text, and intended behavior, with no indication that users may choose another language. The provided policy says to flag language or locale constraints unless they are optional or clearly justified as region-specific.
The argparse description and help messages are all written only in Chinese, which constitutes a language/locale restriction in user-facing natural language. There is no user choice, fallback, or justification indicating that the tool is intentionally limited to a Chinese-speaking context.
No suspicious patterns detected.
