Resume Score for the resume

• Outcome: Review
• Latest audit: May 29, 2026 · 9:33 PM UTC
• Version: 1.0.0

Overview

The skill is a purpose-aligned resume scorer, but it needs Review because its helper script can read local files and automatically writes report data to a fixed desktop path without clear user control.

Install only if you are comfortable providing resume contents to the skill. Use it with pasted text or an explicitly selected resume file, avoid pointing it at unrelated local files, and be aware that running the Python helper directly may create a report at a fixed desktop path containing resume-derived information.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands

Findings (6)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 87% confidence
• Finding: The skill explicitly supports reading user-supplied resume files and references assets that can generate output artifacts, yet it declares no permissions. Missing permission declarations weaken reviewability and can cause users or hosting systems to underestimate file-system access, increasing the chance of unintended local file reads or writes. In this context, resumes contain sensitive personal data, so undeclared file capabilities are more concerning than in a non-sensitive text-processing skill.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 91% confidence
• Finding: The declared purpose is resume scoring, but the implementation description indicates additional behaviors such as writing results to a fixed local path, emitting raw JSON, and exposing extra interfaces. This mismatch reduces transparency and can lead to unintended persistence of sensitive resume data or metadata outside the user's expectation, especially when outputs are written automatically to disk. The fixed-path write behavior makes the issue more dangerous because it can overwrite files or leak private data into predictable locations.

Description-Behavior Mismatch

Medium

• Confidence: 95% confidence
• Finding: The CLI path always writes resume contents and scoring results to a hard-coded desktop file, which creates unintended local data exfiltration and privacy risk for sensitive personal information in resumes. This behavior exceeds the core scoring function and can leak or persist PII without user consent, especially in shared or managed environments.

Context-Inappropriate Capability

Low

• Confidence: 79% confidence
• Finding: The script accepts an arbitrary local file path from the command line and reads it if it looks like a short path-like argument. In isolation this is standard CLI behavior, but in an agent skill context it broadens access from 'score uploaded resume text' to reading any readable local file, which can expose unrelated local data if the surrounding agent passes attacker-influenced paths.

Missing User Warnings

Medium

• Confidence: 84% confidence
• Finding: The skill is designed to ingest resumes, which commonly contain PII such as names, phone numbers, email addresses, location, education, and employment history, but the description provides no privacy notice or handling constraints. Without explicit warning and minimization guidance, users may unknowingly expose sensitive data, and the system may process more personal information than necessary. The risk is heightened here because the skill also contemplates file input and report generation, which can increase retention and propagation of that data.

Missing User Warnings

Medium

• Confidence: 97% confidence
• Finding: Writing a report to a fixed desktop location without warning or consent can disclose sensitive resume data and derived assessments to other local users, backup systems, or endpoint monitoring tools. Because the destination is fixed, users cannot easily prevent or redirect this persistence, making accidental leakage more likely.

VirusTotal

67/67 vendors flagged this skill as clean.

Static analysis

No suspicious patterns detected.