Security audit
Ai Visibility
Security checks across malware telemetry and agentic risk
Overview
This is a markdown-only skill that gives guidance for checking brand visibility across AI platforms and does not include executable code or hidden system access.
Reasonable to install if you want help assessing brand visibility across AI tools. Avoid entering confidential launch plans, private customer data, or sensitive competitive information into visibility checks unless you are comfortable with your agent sending that text to external AI or search services.
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.