ClawHub
Back to skill

Security audit

VDA6.3过程审核辅助

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed VDA 6.3 audit-assistance skill; its trigger wording is somewhat broad, but the artifact does not show hidden, privileged, persistent, or destructive behavior.

Before installing, be aware that generic audit or report requests might invoke this VDA 6.3-specific workflow; confirm that you actually want VDA 6.3 process-audit assistance when it activates. No artifact-backed evidence of malware, exfiltration, destructive behavior, or hidden privilege use was found.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (1)

Vague Triggers

Medium
Confidence
84% confidence
Finding
The trigger conditions are broad enough to activate on generic mentions of audits, standards, report generation, or self-assessment, which can cause the agent to invoke this specialized skill outside its intended scope. That can lead to context hijacking, inappropriate workflow steering, or disclosure/use of the skill's internal guidance when the user did not actually request VDA 6.3 assistance.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.