ClawHub
Back to skill

Security audit

unisound-treatment-process

Security checks across malware telemetry and agentic risk

Overview

This skill is not malicious, but it needs review because it sends medical-record text to a remote model and can save patient-derived text locally.

Install only if the configured model endpoint is approved for your medical data. Use de-identified records, avoid --save-prepared unless you have a secure retention plan, store any output in a protected location, and treat the appkey and generated files as sensitive.

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 (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
95% confidence
Finding
The skill declares no explicit permissions, yet its documented behavior clearly includes reading local files, optionally writing output files, and sending patient-derived medical text to a remote API. This creates a transparency and governance gap: operators may approve or run the skill without realizing it performs network exfiltration of sensitive health data and filesystem access, increasing the risk of unauthorized data handling.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The script sends raw medical-record-derived prompt content to a remote HTTPS LLM endpoint, but the CLI and code provide no explicit consent gate, warning, minimization, or redaction step before transmission. In a medical context, this can expose highly sensitive patient data to external processing, creating privacy, compliance, and data-governance risk even if the endpoint is an internal service.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
When --save-prepared is used, the script writes preprocessed medical text to disk in cleartext without any warning, access control, retention policy, or sanitization. Because the prepared text is derived directly from source records, it may persist sensitive patient information in locations that are later backed up, shared, or left readable by other users/processes.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The generated treatment-process output is written to disk without warning, even though it is patient-derived content and may still contain identifying or sensitive clinical information. This creates local data-exposure risk through filesystem persistence, backups, logs, or accidental sharing of output artifacts.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.