Security audit
Security checks for vulnerabilities and agentic risk
This skill is a local resume-to-job matching helper with no evidence of hidden network access, persistence, credential use, or destructive behavior.
Before using it, treat resume and job materials as sensitive: redact addresses, phone numbers, private employer data, and anything not needed for matching. Expect the bundled script to provide only a keyword-overlap JSON draft; review any rewritten bullets yourself to ensure they do not exaggerate or invent experience.
The declared description promises tailored rewriting, gap quantification, and an interview evidence map, but the analyzed behavior appears limited to simple keyword scoring. This mismatch is dangerous because users and orchestrators may trust the skill to perform richer processing than it actually does, leading to misleading outputs, over-trust, and unsafe automation decisions based on incomplete analysis.
The skill advertises use of a local script and structured file generation, but it does not declare any explicit tool scope or permissions despite detected file read/write capabilities. This weakens least-privilege controls and makes it harder for a host agent or reviewer to bound what the skill may access or modify, increasing the chance of unintended file operations.
An overly broad trigger phrase can cause the skill to activate for generic resume-help requests that may not require this specific workflow. That increases the chance of accidental invocation and unnecessary collection or processing of sensitive employment data in contexts where the user did not intend to use this skill.
A trigger phrase that broadly matches employment-related requests may collide with unrelated job-search assistance workflows. In this skill's context, that can funnel users into sharing resume and job-description content prematurely, creating avoidable privacy exposure and misrouting requests.
The skill explicitly asks for resume text, job descriptions, seniority, region, industry, and portfolio details, all of which can contain sensitive personal and employment information. Without an explicit privacy notice, minimization guidance, or data-handling warning, users may disclose more personal data than necessary without understanding retention, exposure, or redaction risks.
The skill description uses Chinese section headings and bullets alongside English trigger phrases and input/output labels, which can impose a language/locale expectation on users without explicit opt-in. This may conflict with language-choice policy when the skill does not state that users can interact in their preferred language.
This markdown file uses Chinese throughout its substantive instructions and checklist content, which can impose a language requirement on users or reviewers without explicit opt-in. The policy allows locale constraints when clearly documented and justified, but this file does not state that it is intended only for Chinese-speaking users or a region-specific workflow.
The example trigger list mixes specific phrases with broader ones like "tailor my resume" and "job match analysis," without clarifying boundaries or exclusion conditions. In a markdown skill description, this can make activation scope unclear and increase the chance of unintended invocation for general career-assistance requests.
This markdown file presents all operational instructions and validation criteria in Chinese, with no indication that language selection is optional or that the skill is intentionally limited to a Chinese-speaking context. Under the policy rule, forcing a specific language without user opt-in can be a natural-language policy violation.
No suspicious patterns detected.
