ClawHub
Back to skill

Security audit

Lab Result Interpretation

Security checks across malware telemetry and agentic risk

Overview

The skill appears intended to explain lab results, but some included templates overstate diagnoses or treatments and the file-reading CLI is less constrained than its documentation claims.

Review this skill carefully before installing. Use it only for educational lab-result explanations, not diagnosis or treatment decisions, and verify abnormal or critical results with a qualified healthcare professional. If using the CLI, provide only intended lab-report files and avoid passing arbitrary local paths until the file-path validation and dependency pinning are fixed.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Intent-Code Divergence

Medium
Confidence
88% confidence
Finding
The audit explicitly says path traversal protection for the --file parameter is only documented and was not verified in code. If the skill reads attacker-controlled file paths without canonicalization and workspace-boundary checks, an attacker may access unintended local files, which is a real security issue even in a medical lab interpretation context because the tool processes user-supplied input and may run with access to sensitive workspace data.

Unpinned Dependencies

Low
Category
Supply Chain
Content
dataclasses
Confidence
95% confidence
Finding
dataclasses

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.