Security audit
Security checks for vulnerabilities and agentic risk
The skill is mostly a structured-analysis prompt, but it also documents command execution, file writing, external APIs, and API keys without clear limits.
Review before installing. Use it only for local, user-directed structured analysis unless you are comfortable separately approving any command execution, file writes, external API calls, or API-key use. The artifact does not show malicious code, but its higher-impact capabilities need clearer limits.
Documenting command execution for a skill whose stated purpose is analysis materially increases risk because shell or exec access can perform arbitrary system actions unrelated to analysis. In this context, the capability is not narrowly justified, so it could be invoked to read sensitive files, alter the environment, or run harmful commands if an agent trusts the documentation.
Overly broad trigger keywords like analysis, code, data, input, structured, and analyze are likely to match ordinary requests and cause the skill to activate unexpectedly. When combined with the document's broader-than-advertised capabilities, accidental invocation raises the chance of inappropriate tool use, privilege exposure, or user-confusing behavior.
The description states "支持中文交互，无需复杂配置即开即用," and the display name and core messaging center the skill around Chinese interaction, but the document does not clearly offer a user language choice or explain a justified region-specific constraint. This can be a language/locale policy issue because it implies a preferred language behavior without explicit opt-in.
The skill is presented as a structured-analysis tool, but the documentation expands its scope to file handling, external API use, and command execution. That mismatch can cause an agent or user to authorize broader capabilities than expected, increasing the risk of unintended side effects or privilege misuse under the guise of harmless analysis.
The documented ability to call external APIs expands the skill from local analysis into data exfiltration and outbound communication territory. For an analysis skill, that is a meaningful security boundary change because user-provided or locally available data could be sent to third parties without a clearly justified need or disclosed trust model.
L096-L097 state that the skill is Markdown-based and requires no additional API key except explicitly marked external APIs. Later, L193-L195 say that some functions do require platform-specific API keys via environment variables, creating an active documentation contradiction about the skill's operational requirements.
The manifest description and domain table at L007-L008 and L060 present decisions as a supported analysis domain. L114 then says the skill is not suitable for complex decision scenarios requiring human judgment, which conflicts with the broad earlier claim without clarifying the boundary.
No suspicious patterns detected.
