Branding
Security checks across malware telemetry and agentic risk
Overview
This is a simple branding guidance skill with no evidence of code execution, credential access, persistence, or hidden behavior.
This appears safe to install as a branding workflow aid. Users should still independently verify domains, trademarks, social handles, accessibility contrast, and cultural or language checks before using final brand assets publicly.
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
65/65 vendors flagged this skill as clean.