Security audit
Security checks for vulnerabilities and agentic risk
This skill is a local inbox-triage helper that produces reviewable Markdown reports and does not show evidence of hidden network access, message sending, deletion, persistence, or credential use.
Install this if you want a Chinese-language, local inbox/message triage template. Review any input before use because inbox text can contain private information, and only provide an --output path you are comfortable having overwritten. The unused audit code should be cleaned up in a future version, but it did not change the final verdict because it is inactive in the shipped configuration.
声明描述的是邮件/消息分拣助手，核心能力应是把输入通信内容整理成‘回复/等待/跟进/归档’等行动视图，并明确不执行发送或删除。实际代码并未体现任何邮件或消息语义处理、分类逻辑、收件箱工作流集成，甚至没有针对邮件字段、线程、发件人、时间等进行分析。相反，它是一个通用的本地审计/报告生成工具，支持目录审计、CSV 审计、正则模式安全扫描以及 Skill 结构检查。这些能力与‘邮件整理’主目的明显不同，且涉及本地文件系统读取与扫描这一未在描述中体现的资源访问范围。因此描述与代码行为存在实质性不匹配。
The code recursively scans directories, reads many local text files, parses CSV/TSV data, inspects SKILL.md metadata, and performs regex-based secret and private URL discovery. For an inbox-action-board skill, these capabilities are out of scope and create an unintended local content discovery surface that can expose sensitive repository structure, document contents, headings, and potential secrets in generated reports.
The dispatcher enables multiple generic analysis modes such as directory auditing, CSV auditing, pattern scanning, and skill auditing that are unrelated to the stated purpose of organizing inbox items into action categories. This broadens the skill into a general local file inspection tool, increasing the chance that a user or upstream agent can use it to enumerate and summarize arbitrary repository contents and sensitive local data.
The trigger phrases are short and generic enough to overlap with ordinary inbox or productivity requests, which can cause this skill to activate in situations the user did not explicitly intend. Over-broad routing is risky because it may steer unrelated conversations containing emails, messages, or task content into this skill’s processing path, increasing the chance of unnecessary exposure of sensitive communications or incorrect task handling.
This markdown file is entirely written in Chinese and provides no indication that the user may choose another language or that the skill is intentionally limited to a Chinese-language or region-specific context. Under the policy rule for natural-language violations, forcing a specific language without opt-in is reportable.
The skill declares itself as a read-oriented inbox triage tool, but it also documents shell execution and implies file input/output without any explicit permission or allowed-tools declaration. This creates an unsafe trust boundary: a caller or orchestrator may expose shell and file capabilities the user did not expect, enabling unintended local file access or script execution.
The title, category label, summary, and other user-facing fields are written entirely in Chinese, which indicates a fixed language/locale experience. The file does not provide any indication of user choice, multilingual support, or a documented reason for restricting the skill to Chinese.
The entire template, including the title and all section headings, is written only in Chinese and provides no indication that users may choose another language. Under the policy for natural-language constraints, forcing a specific language without user opt-in is a violation unless the locale restriction is clearly documented and justified.
The generated report strings in make_structured_report are written in Chinese, and similar fixed Chinese output appears throughout the script. This imposes a specific language on users without offering a language choice or documenting a justified locale constraint, which matches the policy-violation category.
The directory_report function emits headings and guidance exclusively in Chinese, such as the scan report title and review instructions. Because the script provides no language-selection mechanism, this is a natural-language locale policy issue rather than a functional requirement communicated to the user.
Both csv_report and pattern_report produce user-facing report text entirely in Chinese. The file does not provide an opt-in, locale switch, or documented regional justification, so the skill forces a language choice on all users.
The skill_audit and parse_frontmatter paths return multiple user-visible Chinese messages, including validation errors and audit section text. Since the script exposes these messages directly and lacks any language preference handling, it violates the language/locale policy described in SQP-3.
The manifest presents the skill as an organizer for inbox action views and warns against using it for direct message operations like sending or deleting. While this code does not send or delete messages, it does support writing reports to arbitrary filesystem paths, which is a broader side effect than a simple inbox-view organizer would typically need and is not justified by the manifest wording.
The manifest description contains Chinese and English directives together, implying a fixed bilingual presentation rather than letting the user choose their preferred language or locale. The policy requires flagging language or locale constraints when a skill does not explicitly offer user opt-in or choice.
The file is entirely written in Chinese and includes requirements such as '使用清晰标题', which imply the skill operates in a specific language/locale without documenting user opt-in or offering alternatives. This can violate language/locale policy when a skill constrains output language by default.
This markdown file presents all headings and instructional text in Chinese and does not indicate that the user can choose another language. Per the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale constraint is documented and justified.
This markdown file presents all instructions and expected outcomes exclusively in Chinese. Under the policy for natural-language violations, forcing a specific language without user opt-in can be a locale/language policy issue when no alternative or choice is provided.
No suspicious patterns detected.
