Security audit
Security checks for vulnerabilities and agentic risk
This skill mainly generates Chinese weekly-review drafts from user-provided input, with no network access, persistence, or hidden execution, though its bundled script contains inactive audit modes that should stay disabled unless reviewed.
Install this if you want a Chinese-language weekly review drafting helper and are comfortable providing the work-log content it will summarize. Keep inputs limited to the notes you intend to include, avoid sensitive personal or business data unless appropriate, and be cautious using --output because it writes to the path you provide. The extra audit modes in the script should not be enabled without a separate review and clearer documentation.
声明描述的是一个面向周报/复盘的摘要技能，但代码并不专门实现“本周完成、阻塞、经验教训、下周计划”的业务逻辑。相反，它通过 spec['mode'] 支持多种泛化审计与扫描能力，包括目录审计、CSV 审计、模式扫描和 skill 合规检查。这些能力明显超出并偏离了周报汇总用途，且会访问本地文件系统中的目录、文本文件和表格数据进行扫描分析。虽然其中的 structured_brief 模式可以生成结构化摘要，表面上可部分用于周报，但整体代码的主要能力是通用报告/审计框架，而非单一的周报总结工具，因此描述与实际行为存在实质性不匹配。
The script’s implemented capabilities materially exceed the stated weekly-review summarization purpose by supporting directory audits, CSV audits, pattern scanning, and skill/package compliance checks. This kind of scope mismatch is dangerous because a user invoking a benign-seeming productivity skill could unintentionally expose local repository contents, filenames, headings, and matched sensitive snippets through generated reports.
The pattern scanning logic searches arbitrary files/directories for secrets, private URLs, and dangerous shell patterns, then includes matched snippets in output. In the context of a weekly review skill, this is an unjustified data-inspection feature that can leak sensitive content from local files into console output or saved reports.
The README lists invocation examples like '根据本周工作日志生成周报' and '帮我做个人 weekly review' without defining explicit trigger boundaries, exclusions, or a narrow invocation context. These phrases are common work-assistance requests and could overlap with ordinary conversation, making unintended activation more likely.
The skill advertises itself as a reporting helper but also documents shell execution and implies read/write interaction with local resources, while declaring no explicit tool scope or permission boundaries. This creates an ambient-authority problem: a caller or host may expose broader capabilities than intended, enabling file access or command execution beyond a simple weekly review workflow.
The entire example output is written in Chinese and the title indicates a fixed Chinese reporting format. There is no visible indication that the user can choose another language or that the locale restriction is justified, which can violate language/locale policy requirements.
Multiple hardcoded user-facing strings in the report generator are in Chinese, causing the skill to emit a fixed language regardless of user preference. The policy allows locale constraints only when the user is given a choice or when the restriction is clearly documented and justified, neither of which appears in this file.
Directory, CSV, and pattern report functions all generate user-visible headings and guidance in Chinese only. This enforces a specific locale across multiple modes without any visible mechanism for user choice or a documented justification for the restriction.
The skill audit functionality inspects package structure and parses SKILL.md metadata, which is unrelated to weekly review generation. While not directly destructive, it enables filesystem enumeration and content inspection under a misleading skill identity, increasing the risk of unintended disclosure and trust abuse.
The frontmatter parser and skill audit renderer use Chinese user-facing error and status text, again imposing a fixed language on users. There is no evidence in this file of opt-in language selection or a clearly documented regional limitation.
The entire smoke test file is written in Chinese and does not indicate that language selection is optional or limited to a justified region-specific context. Under the policy rule for natural-language violations, this can amount to forcing a specific language without user opt-in.
All user-facing documentation in this file is presented in Chinese, and the skill does not indicate that users may choose another language or that the Chinese-only presentation is required for a specific region or compliance context. This can be a language/locale policy issue when a skill effectively forces one language without opt-in.
The file is entirely framed in Chinese and describes the expected output structure without indicating that users may choose another language or locale. Under the policy rule for language/locale choice, this can be interpreted as a natural-language constraint that defaults to a specific language without opt-in.
The title, category label, summary, input hint, sections, and most examples are in Chinese, which suggests the skill is primarily configured for Chinese output or interaction. However, the file does not explicitly state that the skill is region-specific or that users may choose their preferred language, so this can be read as a locale/language default without clear opt-in.
The template headings and instructions are entirely in Chinese, which implies a fixed language requirement for the skill output. Under the policy, language constraints should either be optional for the user or clearly justified as region-specific; neither is stated here.
This code performs a filesystem write by creating parent directories and writing rendered output to the user-supplied path. While the CLI argument name suggests output behavior, there is no confirmation prompt or inline warning comment/docstring near the write operation explaining that it will modify files and potentially overwrite existing content.
No suspicious patterns detected.
