ClawHub
Back to skill

Security audit

voc-conversation

Security checks across malware telemetry and agentic risk

Overview

This skill analyzes user-provided customer service conversations locally and tells the agent to redact sensitive details before reporting quotes.

Install only if you are authorized to process the customer conversations you provide. Prefer minimizing or redacting sensitive fields before analysis, and review generated quotes to ensure they cannot identify customers before sharing reports.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The README explicitly describes processing customer service conversation exports and producing 'real verbatim quotes', which commonly contain personal data, account details, contact information, or complaint narratives tied to identifiable individuals. Although it mentions local processing and automatic desensitization later, it does not clearly instruct users to ensure lawful authorization, minimize sensitive data, or verify that quoted output cannot re-identify individuals, creating a real privacy/compliance risk.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.