Security audit
Security checks for vulnerabilities and agentic risk
This is a coherent Chinese novel-writing skill that writes story files and continuity notes locally, with a scoped cleanup option users should handle carefully.
Install this if you want a Chinese long-form novel workflow that stores generated chapters and continuity notes in local markdown files. Avoid using --clean unless you are comfortable deleting prior generated output and resetting the story memory files.
Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).
if [ "$CLEAN" = true ]; then
log_step "清除旧的输出文件..."
rm -rf "$OUTPUT_DIR"/*.md 2>/dev/null || true
log_step "重置记忆文件..."
The natural-language description presents the skill as a Chinese-language novel generator and all usage examples and instructions are in Chinese. There is no indication that users may choose another language or that the Chinese-only behavior is an intentional, documented locale constraint.
The README states that the skill will output chapter files and persist memory and error logs, but it does not clearly warn users that local files will be created and retained over time. In an agent-skill context, silent persistence can expose sensitive prompts, story ideas, personal data, or workspace contents to unintended retention and later reuse, especially if the agent runs in a shared repository or syncs files to version control.
The activation description is very broad and matches generic creative-writing requests, which increases the chance the skill will auto-trigger in situations where the user did not specifically want file-writing, memory management, or multi-step workflow behavior. In this skill, that matters because activation can lead to reading and writing project files such as .learnings/ and output/, expanding the skill’s operational scope beyond simple text generation.
The metadata fixes the skill language to zh-CN, and the document consistently instructs Chinese-language operation without indicating that users may choose another language. This can violate language-choice policy when the skill is used in broader contexts not explicitly limited to a China-specific compliance or regional workflow.
The skill instructs use of an initialization script that clears prior .learnings/ records, but it does not require an explicit warning, backup, or user confirmation before destructive action. This creates a real risk of accidental data loss, especially because .learnings/ is the persistence mechanism for story continuity and may contain valuable prior work.
This markdown template is entirely written in Chinese and defines section headings and field values in Chinese, which implicitly constrains downstream skill use to a specific language. The file does not offer an opt-in, alternative locale, or justification that the skill is intended only for a Chinese-language context.
The file title and instructions are entirely in Chinese and direct the model to auto-complete prompts in that language. This imposes a specific language/locale by default, and the template does not provide any opt-in or alternative language selection for users.
The --clean path performs destructive actions by deleting prior output markdown files and overwriting multiple memory files with no confirmation, backup, or dry-run safeguard. In an agent-integrated workflow, accidental invocation or misuse can cause irreversible loss of user-authored content and project state.
SQP-3 applies to all file types and includes language or locale policy violations. This markdown file presents all instructions and content exclusively in Chinese, and nowhere indicates that the user may choose another language or that the skill is intentionally limited to a Chinese-language context.
SQP-3 applies to all file types and covers natural-language policy violations such as forcing a specific language without user opt-in. This document presents all guidance in Chinese with no indication that users may choose another language or locale, which can amount to an implicit language constraint.
All comments, usage text, logs, and prompts are presented only in Chinese, which imposes a specific language on users without any opt-in or documented locale limitation. Under the language/locale policy, a skill should either offer a language choice or clearly justify being region- or audience-specific.
No suspicious patterns detected.