ClawHub
Back to skill

Security audit

unisound-med-mental-health

Security checks across malware telemetry and agentic risk

Overview

This mental-health Q&A skill is a disclosed command-line wrapper around a remote medical LLM API, but users should treat all prompts and saved outputs as sensitive health data.

Install only if you are comfortable sending mental-health questions to the configured remote medical LLM endpoint. De-identify real patient or personal data first, avoid putting secrets in prompts, use --dry-run or --text-only when appropriate, and protect any --output files or terminal logs because they may contain the full question and metadata.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding
The skill documentation describes capabilities that read local files/stdin, write output files, and send prompts to a remote medical-model API, but it does not declare corresponding permissions. This creates a transparency and policy-enforcement gap: integrators may deploy it without realizing it can access local data and transmit sensitive mental-health content over the network, increasing the risk of data leakage or misuse.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The code sends raw mental-health questions and associated metadata to a remote third-party API endpoint via `urlopen` without any built-in consent, disclosure, minimization, or redaction. In this skill's context, prompts may contain highly sensitive health information, so silent transmission materially increases privacy, compliance, and data-governance risk.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.