ClawHub
Back to skill

Security audit

短视频带货脚本生成器

Security checks across malware telemetry and agentic risk

Overview

The artifact appears coherent and purpose-aligned, with no validated scanner findings or artifact-backed signs of hidden behavior.

Install or use this only if you trust the ClawHub/OpenClaw tooling and are comfortable configuring registry, GitHub, Convex, or OpenAI credentials where needed. Review the moderator and maintainer skills carefully before invoking them because they can affect public listings, users, roles, or PR state when run with authorized credentials.

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

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.