ClawHub
Back to skill

Security audit

ZenVFX CLI

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed ZenVFX helper skill whose main risks are expected for managing cloud video projects: installing a CLI, using a token, editing canvases, uploading files, and deleting ZenVFX files.

Install this only if you trust the ZenVFX/Tencent CLI and intend to give it access to your ZenVFX account token. Before using destructive commands such as file:rm or node/edge removal, verify the target path or IDs, and avoid uploading sensitive local files unless they belong in the ZenVFX project.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
87% confidence
Finding
The skill documents a destructive deletion command, `file:rm --path <p>`, without requiring confirmation, dry-run behavior, or explicit user authorization. In an agent context, broad triggers and procedural guidance can lead to accidental or over-broad deletion of project files or directories, especially if path values are inferred from ambiguous user requests.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.generated_source_template_injection

User-controlled placeholder is embedded directly into generated source code.

Critical
Code
suspicious.generated_source_template_injection
Location
SKILL.md:504