Security audit
Security checks for vulnerabilities and agentic risk
The skill is a coherent resume builder, but it under-discloses that it stores sensitive personal profile data and may fetch job descriptions from URLs.
Install only if you are comfortable with the agent storing your resume profile in the workspace and potentially fetching user-provided job-description URLs. Prefer pasted job descriptions, review resume-profile.md for sensitive fields, and delete or redact it when no longer needed.
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
</head>
<body>
<div class="resume">
<div class="header"> <!-- Name, target, contact -->
<div class="section edu"> <!-- Education (brief, near top) -->
<div class="section"> <!-- 3 Core Advantages grid -->
<div class="section"> <!-- Key Projects -->
The manifest description says to use the skill when the user wants to create a resume 'from scratch' or says 'make me a resume,' which are broad natural-language phrases that could overlap with ordinary conversation. The file does not provide exclusion conditions or tighter trigger constraints to limit when those phrases should activate the skill.
The manifest claims there are no network requests, but the skill later explicitly accepts a JD via URL using web_fetch. This mismatch undermines the security boundary communicated to users and reviewers, and can cause the agent to access external content when the documentation says it will not. In this context, the skill also handles highly sensitive resume data, so inaccurate disclosure increases risk.
The security text frames headless browser usage as limited to local PDF export, but the skill also authorizes unrelated network retrieval through URL input. That creates a misleading trust signal: operators may approve the skill believing browser use is isolated and offline, while actual behavior includes external data access. The danger is amplified because fetched content is untrusted and may influence generated output or expose request metadata.
The skill collects and stores sensitive personal data including name, phone, email, birth date, location, gender, education, and work history, and saves it to a persistent workspace file without an explicit privacy warning or consent step. This creates a real privacy and security risk because users may disclose regulated or high-value PII without understanding retention, storage location, or downstream exposure. The context makes this more dangerous because resume data is rich identity information and the workflow encourages persistence across sessions.
The sample HTML sets lang="zh-CN" as the default document language, which imposes a specific locale in a general-purpose resume template. Although later localization guidance mentions both Chinese and English resumes, this line itself does not present the locale as optional or user-selectable.
No suspicious patterns detected.