Security audit
Security checks for vulnerabilities and agentic risk
This appears to be an offline medical-device classification helper, but its package metadata requests broader agent powers than the tool needs and its integrity metadata is inconsistent.
Review this before installing because the package should narrow its allowed tools and republish accurate integrity hashes. If used, grant only the minimum needed to run the local Python CLI, avoid Write/WebFetch unless separately justified, and verify all classification results against current official regulatory sources or qualified counsel.
The document title and introductory guidance are written entirely in Chinese and direct readers to other materials without indicating any language choice or opt-in. This can violate language/locale policy when the skill is expected to be generally usable across users or agents with different language preferences.
The trigger list includes broad generic terms such as “医疗器械”, “医械合规”, and “判定器”, which can cause the skill to activate outside its narrowly intended use. In an agent environment, overbroad invocation can route unrelated medical-regulatory queries into a deterministic screener, increasing the chance of incorrect compliance guidance or unintended tool use.
The natural-language description and usage context are presented entirely in Chinese, which imposes a specific language on users without any stated opt-in or alternative language support. The policy allows locale constraints when explicitly justified, but this file does not document such a limitation or offer user choice.
The file’s docstrings, CLI help text, error messages, warnings, and output disclaimer are all written in Chinese, and there is no option for the user to select another language. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale restriction is explicitly documented and justified.
The document is explicitly split into English and Chinese sections, but there is no indication that language output is optional or user-selectable. Under the stated policy, forcing specific languages without opt-in can be a natural-language locale violation unless the restriction is clearly justified.
The changelog explicitly states "Bilingual rewrite (full EN+ZH, English first)", which imposes a presentation order preference rather than offering a user-selected language or locale. The file includes both English and Chinese content, but does not state that language choice is user-configurable or optional.
The visible text in the asset is primarily Chinese, which imposes a specific language choice on users. For a non-region-scoped skill artifact, this can violate language/locale policy when no alternative language option or justification is provided.
Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.
"rc": 2, "error_code": e.code}, ensure_ascii=False))
sys.exit(0 if ok else 2)
args = {k: getattr(ns, k) for k in names}
if not any(v not in (None, "") for v in args.values()):
p.print_help()
sys.exit(0)
No suspicious patterns detected.