Security audit
Brand Identity
Security checks across malware telemetry and agentic risk
Overview
This appears to be an instruction-only brand identity skill with no evidence of hidden execution, credential access, persistence, or data exfiltration.
This skill looks safe to install for brand identity work. As with any creative assistant skill, review outputs before using them publicly, especially for legal names, trademarks, claims, or brand promises.
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
63/63 vendors flagged this skill as clean.