Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed Chinese K12 learning-profile system with sensitive student-memory features, but its artifacts consistently require consent, identity checks, minimization, and user controls.
Install only in a Chinese K12 deployment that can enforce per-student isolation, guardian consent where required, field-level sharing controls, and deletion/export requests. Do not use it outside that locale until crisis contacts, curriculum assumptions, and minor-consent rules are localized.
该代码块仅涉及数据模式验证，属于开发/测试辅助逻辑。虽然它间接反映了某种“档案数据结构”的存在，但没有任何持久化存储、检索、更新、删除、导出、权限判断、触发条件处理、身份核验或敏感字段单独授权控制等核心功能。声明强调的是一个处理敏感未成年人学习档案的业务技能，而代码的实际行为只是校验 JSON 是否符合 schema，主用途与能力范围都明显不一致，因此应判定为描述与行为不匹配。
The profile stores understanding-depth assessment fields even though the manifest says this skill does not perform understanding verification and should only receive such data via handoff. In practice, examples often drive implementation; this can cause the skill to over-collect sensitive educational profiling data about a minor, violating purpose limitation and weakening consent boundaries between skills.
The crisis-handling instruction hardcodes '中国大陆即时危险为 110/120' in the same sentence that discusses asking for the user's region if uncertain. This creates a locale-specific policy issue because the skill embeds a specific regional emergency convention rather than consistently requiring locale confirmation before presenting contacts.
Suspicious Unicode normalization or mixed-script content
The entire skill file is written in Chinese and presents instructional content only in that language, with no indication that users may choose another language or that the skill is intentionally limited to Chinese-speaking users. Under the stated policy, forcing a specific language without user opt-in is a natural-language policy violation.
The entire template, including required opening lines and scripted prompts, is written as mandatory Chinese interaction for the user-facing experience. There is no indication that the user may choose another language, nor is there a documented region-specific justification for enforcing this locale.
The trigger list explicitly notes that these phrases are common in ordinary tutoring, yet it still operationalizes them as archive-update prompts. In a skill handling sensitive minors' longitudinal learning records, vague conversational triggers can cause unintended collection or mutation of profile data during routine study chat, especially if downstream systems or agents mis-handle the 'confirm before write' requirement.
The entire skill documentation is written in Chinese and does not indicate that language selection is optional or that the skill is intentionally limited to a Chinese-speaking context. Under the policy, forcing a specific language without user opt-in is a natural-language locale violation.
This JSON schema is a manifest/config-type file, so natural-language policy checks apply. The top-level description is written only in Chinese and the schema continues to define user-facing labels and values exclusively in Chinese, with no note that the skill is China-specific or that users can opt into this locale, which can violate language/locale choice policy.
The example profile enables reminder-related consent and stores consecutive-learning-day tracking even though the skill description says it does not send reminders. In a minor-focused learning archive, this creates scope creep: downstream implementers may infer that reminder behavior and habit surveillance are permitted, leading to collection or use of data beyond the declared purpose and consent model.
This JSON example is written entirely in Chinese for field values and user-facing text, including consent labels, goals, comments, and strategy descriptions. For a general-purpose profile example, that effectively enforces a specific language/locale without any visible opt-in or documentation that the schema is intentionally region-specific.
The example includes oral/pronunciation growth tracking that is not clearly justified by the manifest's stated archive scope. For a system handling minors' longitudinal records, undocumented expansion into speech-performance profiling increases privacy risk and may enable broader behavioral or biometric-adjacent tracking than users and guardians were told to expect.
The file is entirely written as mandatory protocol text in Chinese and gives prescriptive requirements such as required labels and review steps, but it does not state that Chinese is optional or that users may choose another language. Under the language/locale policy rule, forcing a specific language without opt-in is a natural-language policy concern.
This markdown file is entirely written in Chinese and provides operational guidance for skills, but it does not indicate that users may choose another language or locale. Under the policy for natural-language violations, forcing a specific language without user opt-in is a reportable issue unless the locale restriction is clearly justified and documented.
This JSON manifest/config file contains natural-language policy-relevant text entirely in Chinese, including operational notes that govern how the skill copy is used. Because the file does not indicate that Chinese is optional, user-selected, or justified as a region-specific constraint, it presents a language/locale policy issue under the rule for natural-language policy violations.
The main description field defines the protocol in Chinese and serves as the primary human-readable explanation for the schema. With no accompanying locale choice, English fallback, or region-specific justification, this is a natural-language locale constraint embedded in the file.
The file is entirely in Chinese and provides no mechanism to detect or honor the student's preferred language. In a skill handling sensitive minors' education records and consent-dependent actions, forcing one language can undermine informed consent, identity/authorization clarification, and accurate understanding of restrictions or data-handling choices.
The file hard-codes a China-mainland K12 context and explicitly says the system should not be used elsewhere unless operators first localize emergency contacts, educational standards, and legal consent rules. In a skill handling sensitive minor data and crisis-related facts, defaulting users into the wrong jurisdiction without an explicit user/tenant choice can lead to unsafe crisis guidance, incorrect parental-consent handling, and privacy-law noncompliance.
Lines L004-L008 define the repository-wide vocabulary exclusively in Chinese and require all skills and schemas to use it as the single source of truth. This is a natural-language locale constraint, and the file does not provide any user opt-in, multilingual alternative, or explicit justification that the skill is intentionally limited to a Chinese-language context.
The manifest emphasizes that this skill is a storage/authorization layer and that ordinary Q&A does not read the archive unless the student asks in the current turn. The milestone comment '我开始认识你了' presents the system as generally remembering/using the student profile, which conflicts with the documented limited-read behavior and could mislead developers about intended use.
No suspicious patterns detected.
