Security audit
Security checks for vulnerabilities and agentic risk
The skill’s review workflow is understandable, but it tells the agent to run a local Python script that is not included in the reviewed package.
Install only if you trust the local serve.py that this skill expects to run, because that executable is not present in the reviewed artifact. Prefer a package that includes the server script and clearer instructions requiring explicit user approval before opening the review browser.
This description defines an activation condition that could cause unintended invocation because many common agent tasks involve generating Markdown and 'needs human review' is subjective. Although L10 gives an explicit command, the broad 'Use whenever' language still creates ambiguity about when the skill should auto-activate versus wait for user intent.
The phrase 'needs targeted human feedback before iterating' is ambiguous because it does not specify who decides that need or what types of Markdown documents qualify beyond examples. This can lead to over-invocation in ordinary drafting workflows.
This natural-language instruction implies a fixed language for the user-facing interface, which can violate language/locale expectations when no opt-in or justification is given. The file does not state that the skill is intended only for Chinese-speaking users or offer an alternative label.
