Security audit
Security checks for vulnerabilities and agentic risk
This is a local regression-story/reporting skill with no evidence of exfiltration, persistence, or deceptive execution, though it includes broader dormant audit code and is Chinese-localized.
Use this skill only on issue histories or test-planning files you intend to share with the local agent. Prefer dry-run/stdout or a deliberate output path, because the script can write the generated report to a chosen file. Be aware that the package is Chinese-first and contains inactive audit routines that should not be used on broad private directories unless you intentionally re-scope it for auditing.
代码的主功能与声明描述不一致。声明强调的是测试/QA 场景下，依据历史问题生成回归测试故事、风险等级和优先级；但实际代码没有看到历史问题建模、测试故事生成、风险分级算法、优先级计算等专门逻辑。相反，它实现的是一个由 spec.json 驱动的通用报告生成器/审计器，可扫描目录、解析 CSV、匹配安全模式、检查 Skill 文件完整性和 frontmatter。虽然这些能力可被用于 QA 或审阅辅助，但其主要行为明显更广泛且不同于声明的核心用途，因此属于实质性描述-行为不匹配。
The script implements broad directory auditing and generic pattern scanning capabilities that are materially outside the stated purpose of generating regression stories from historical issues. In an agent setting, this expands the skill’s effective authority to inspect arbitrary local content and produce security-oriented reports, creating unnecessary data exposure and capability creep.
The skill can inspect other skills’ package structure and parse SKILL.md frontmatter, which is unrelated to regression-story generation and enables cross-skill reconnaissance. While not directly executing code, this broadens the tool into a meta-auditor and may reveal internal project layout or metadata that users did not intend this skill to access.
The entire skill documentation is written in Chinese and does not indicate that users may choose another language or locale. Under the policy rule, forcing a specific language without user opt-in is a natural-language policy concern unless the locale restriction is explicitly justified.
The skill advertises executable behavior via python3 .../scripts/run.py and references local resource files, but it declares no explicit tool scope such as shell, file read, or file write permissions. This creates a capability transparency problem: an orchestrator or reviewer cannot reliably constrain what the skill may attempt, increasing the chance of unintended file access or command execution.
The file is entirely written as a Chinese-only example and includes no indication that users may choose another language or locale. Under the policy, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale constraint is clearly documented and justified.
The entire example output is written in Chinese and presents the skill's output format in a single fixed language, with no indication that users can choose another language or that the skill is limited to a Chinese-language context. This can violate language/locale policy when a skill implicitly forces one locale without user opt-in.
The title, category label, summary, and input hint are all presented only in Chinese, indicating a language-specific skill experience. The file does not offer a user language choice or document that the skill is intentionally limited to a Chinese-speaking context, which creates a locale policy concern.
Multiple user-facing report strings are emitted in Chinese, and the script does not provide an option for the user to select another language or opt in to this locale. This is a natural-language policy concern because it forces a specific language in generated output across normal execution paths.
The regex-based scanning for secrets, dangerous shell patterns, and private URLs is not justified by the declared regression-story workflow and gives the skill a security-scanning role it was not supposed to have. This can surface sensitive material from arbitrary files and encourages use of the skill on broader repositories than necessary, increasing confidentiality risk.
The entire template, including headings and instructions, is written only in Chinese and does not indicate that users may choose another language. This can violate language/locale policy when a skill implicitly forces a specific language without user opt-in or documented justification.
This markdown file contains user-facing natural-language instructions only in Chinese, and there is no indication that the skill is region-specific or that users may choose another language. Under the language/locale policy rule, forcing a specific language without opt-in can be a policy violation.
No suspicious patterns detected.
