ClawHub
Back to skill

Security audit

medical-record-structuring

Security checks across malware telemetry and agentic risk

Overview

This skill locally converts user-provided Chinese medical notes into structured FHIR-style JSON, with no evidence of network upload, hidden persistence, or destructive behavior.

Install only if you intend to process medical records locally. Keep generated JSON and Markdown previews in secured locations, avoid pasting real patient data into shared terminals or logs, and review outputs manually because this is an extraction tool, not clinical advice.

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 instructs the agent to invoke local scripts that read input records and write structured outputs to files (for example via stdin, input paths, /tmp outputs, batch directories, and generated JSON bundles), but the metadata declares no permissions. That mismatch is a real security issue because it obscures the skill's filesystem access requirements, weakening policy enforcement and informed approval, especially given the highly sensitive medical data being processed.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The preview output prints raw FHIR resource content derived from medical records, which can include protected health information such as names, identifiers, diagnoses, and procedures. In a healthcare-focused skill, rendering this data without any warning, redaction, or privacy guardrails increases the risk of accidental disclosure through terminals, logs, screenshots, or copied Markdown.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.