Model Benchmark
Security checks across malware telemetry and agentic risk
Overview
This is a markdown-only benchmarking guide with normal cautions around provider API keys and model configuration changes.
Safe to install as a benchmarking guide. Before using it, review any models.json changes, prefer dedicated low-privilege provider API keys, and be aware that benchmark calls may consume provider quota or incur cost.
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.