Security audit
Security checks for vulnerabilities and agentic risk
The skill is not clearly malicious, but it requests write, shell, and web access while telling users it is a pure/no-network documentation checker.
Review this skill before installing. It appears to be a PMCF report checker, but install only if you are comfortable granting it shell, write, and web-fetch capabilities, or ask the publisher to remove unused permissions and add clear privacy/write-scope guidance before using it with confidential medical-device reports.
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
# pmcf-evaluation-report-check · PMCF 评价报告完整性检查器
[](https://github.com/zhaoxinghua09-cell/lgd-theory)
The metadata declares file-writing capability via allowed-tools: ... Write ..., but the user-facing body presents the skill as a 'pure documentation tool' and does not warn that local files may be modified. This mismatch can cause users to supply sensitive local content or run the skill under the false assumption that it is read-only, increasing the risk of unintended overwrites or data alteration.
The metadata permits network access (WebFetch) and the risk rationale mentions API use, but the visible description says it is a zero-dependency, pure documentation checker without a clear privacy or transmission warning. Users may therefore provide regulated medical/compliance report content without understanding that data could be sent externally, creating confidentiality and compliance exposure.
The file presents core usage and licensing information in both English and Chinese, with substantial sections only in Chinese, but it does not state that the skill is intended specifically for Chinese-speaking users or offer an alternate language choice. Under the policy rule, forcing or assuming a language/locale without opt-in can be a natural-language policy concern.
No suspicious patterns detected.