Jobautopilot Submitter

• Outcome: Review
• Latest audit: May 28, 2026 at 2:08 PM UTC
• Version:
  1.3.2

Overview

This job-application automation skill appears purpose-aligned but needs Review because it can handle sensitive resume data, upload files, and answer application questions without clear user confirmation or scoping.

Install only if you are comfortable letting it read configured resume/application files and interact with job application forms. Before use, restrict resume and upload directories, review every file path before upload, and provide explicit answers for language or demographic-style questions instead of relying on defaults.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (4)

Intent-Code Divergence

Medium

• Confidence: 91% confidence
• Finding: The document states that personal data comes exclusively from environment variables, but later instructs the agent to parse tailored resume `.docx` files and extract personal and application data. This inconsistency matters because it broadens the data sources being accessed and weakens the user's ability to reason about what sensitive local files the skill may read.

Intent-Code Divergence

Medium

• Confidence: 90% confidence
• Finding: The safety rules claim all personal data comes from environment variables, yet the workflow separately instructs runtime extraction of work history, education, and contact details from resume files. Conflicting rules around sensitive data sources can lead reviewers or operators to underestimate file access and can enable overcollection of personal information during execution.

Missing User Warnings

Medium

• Confidence: 86% confidence
• Finding: The template explicitly supports file uploads as part of an agent-generated script but provides no built-in user confirmation, allowlist, or path validation before submission. In a job application automation context, this can cause unintended disclosure of sensitive local files or submission of the wrong document if the runtime generation logic is influenced by page content, bad matching, or upstream data errors.

Natural-Language Policy Violations

Medium

• Confidence: 94% confidence
• Finding: The function hard-codes selection of a Chinese/Mandarin language option and silently falls back to "Other" if no exact match is found, without any evidence of user input or consent. In a job-application automation skill, this can cause the agent to submit false self-reported demographic or language information, creating integrity, compliance, and user-harm risks even though it is not a code-execution issue.

VirusTotal

VirusTotal engine telemetry is currently stale for this artifact.