Evaluation Benchmark
Security checks across malware telemetry and agentic risk
Overview
This is a small instruction-only skill for agent evaluation and benchmark planning, with no code execution or hidden access requests.
Install this if you want a lightweight Chinese-language assistant for agent evaluation and benchmark planning. Use explicit evaluation or benchmark wording when invoking it, and avoid sharing private benchmark data or proprietary model outputs unless you are comfortable giving that context to the agent.
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.