Security audit
benchmarking
Security checks across malware telemetry and agentic risk
Overview
This is a plain-text benchmarking skill that guides users in designing, running, and scoring model evaluations without hidden code or privileged behavior.
This skill is appropriate to install if you want help designing or running model benchmarks. When using it, review any benchmark tasks before sharing them with external model providers, because real-work evaluations may include private prompts, operational context, or proprietary scoring criteria.
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
64/64 vendors flagged this skill as clean.
Static analysis
No suspicious patterns detected.