unisound-med-treat

• Outcome: Pass
• Latest audit: May 28, 2026, 5:44 AM
• Version:
  1.0.0

Overview

This skill is a disclosed medical LLM wrapper that sends user-provided questions to a configured API, so it is usable but requires careful handling of patient data and the API key.

Install only if you are allowed to send the selected medical case text to the configured Hivoice/Unisound API. De-identify real patient data, protect the app key, avoid overriding --api-url to an untrusted endpoint, and verify publisher provenance if using it in a clinical or production workflow.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• 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
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The skill sends raw user-supplied medical questions, which may contain sensitive health information, to a remote third-party API endpoint without any explicit consent notice, privacy warning, minimization, or redaction step in the code path. In a medical context this is materially risky because callers may unknowingly transmit protected or highly sensitive personal data outside their local environment.

VirusTotal

VirusTotal findings are pending for this skill version.