performance-benchmark
Security checks across malware telemetry and agentic risk
Overview
This is an instruction-only performance benchmarking skill with expected guidance about running user-chosen benchmark tools and no hidden code or credential use.
Before installing, note that this skill does not supply its own benchmark CLI; you or your agent must choose tools such as k6, autocannon, or hyperfine. Only benchmark systems you are authorized to test, and review .benchmarks/ before committing because reports may include URLs, endpoint names, or performance details.
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.
