ClawHub
Back to skill

Security audit

unisound-pre-existing-review

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed medical-claims review helper that sends user-provided case text to a configured model endpoint and only writes files when the user asks it to.

Install only if your workflow is allowed to send redacted medical or claims text to the configured model service. Redact personal identifiers before use, be aware that normal JSON output includes the question text, and use --output only in locations approved for sensitive case material.

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 (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
85% confidence
Finding
The skill documentation indicates capabilities to read files, write output, and make network requests, but no explicit permissions are declared. This creates a security and governance gap: callers and reviewers cannot accurately assess what the skill may access or transmit, which is especially sensitive here because the skill processes medical/claims material and sends prompts to an internal model endpoint. The context makes this more dangerous because the data may include OCRed records and health information, so undeclared network and file capabilities increase confidentiality and compliance risk.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill sends full medical case text and claim materials to a remote LLM API, which likely contains sensitive personal and health information. Even though the endpoint is described as internal, the code provides no consent prompt, minimization, redaction, or disclosure to the operator, so protected data may be transmitted unexpectedly and logged or retained by downstream services.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill writes output objects to disk that include the original question text and model answer, which in this medical-review context can contain sensitive health and claims data. Because this happens without any warning, sanitization, or secure storage controls, users may unintentionally create local files containing regulated personal information.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.