ClawHub

Class A 品牌化身孵化器

Security checks across malware telemetry and agentic risk

Overview

This is a prompt-only brand avatar generator with no code execution or credential access, though its official-sounding identity language should be treated as creative branding only.

Safe to install for creative brand/persona work. Treat generated DID, identity verification, and legal-coordinate language as fictional or internal branding text, and verify any real legal, trademark, or official registration separately.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (3)

Vague Triggers

Medium
Confidence
84% confidence
Finding
The README tells users to 'wake up this SKILL' and provide only a brand name, with no narrow trigger conditions, input bounds, or safety checks. Overly broad invocation guidance can cause the skill to activate in unintended contexts and process ambiguous user input, increasing the chance of misuse, prompt injection carry-through, or generation of misleading identity-like outputs without sufficient validation.

Natural-Language Policy Violations

Medium
Confidence
95% confidence
Finding
The skill metadata and description are written only in Chinese, which creates a locale lock that can force Chinese-language interaction without user choice. This is not a code-execution risk, but it can cause user confusion, inaccessible behavior, and reduced operator control when the user expects another language.

Natural-Language Policy Violations

Medium
Confidence
98% confidence
Finding
The instruction text directly addresses the agent in Chinese and mandates a specific interaction pattern without offering language choice. In context, this makes the skill more rigid and can override user preferences, which is a prompt-safety and usability issue even though the content itself is not overtly malicious.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal