ClawHub
Back to skill

Security audit

unisound-health-record-management

Security checks across malware telemetry and agentic risk

Overview

This skill has a legitimate health-record purpose, but it sends sensitive medical text to a remote LLM while its privacy controls are under-scoped and partly overstated.

Review this before installing or running with real patient data. Use only de-identified inputs, verify that the configured LLM endpoint is approved for your organization, avoid --output unless you can protect the generated file, and do not rely on the skill itself to remove direct identifiers.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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 advertises executable usage that reads local files, writes output files, and sends sensitive health-record content over the network to an external API, yet it declares no explicit permissions or trust boundaries. In a healthcare context, this omission is dangerous because operators may not realize the skill has data-access and exfiltration capability, increasing the risk of unauthorized PHI handling and policy bypass.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill processes highly sensitive medical record content and sends the full unstructured health text to a remote LLM API, but the CLI gives no explicit privacy warning, consent step, or transmission safeguard beyond a bearer token. In a healthcare context, this creates material risk of unauthorized disclosure, policy noncompliance, or accidental transfer of personally identifiable and protected health information to an external service, especially since the base URL is user-configurable and timeout defaults to indefinite waiting.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.