Security audit
对抗性审查
Security checks across malware telemetry and agentic risk
Overview
This is a documentation-only skill that guides agents to use adversarial review for important decisions, with no hidden code or automatic system access.
Installers should understand that this skill may cause the agent to share selected artifacts and contracts with reviewer subagents or other models during important decisions. That behavior is disclosed and bounded, but users should avoid including secrets in review artifacts unless they intentionally want that data reviewed.
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
63/63 vendors flagged this skill as clean.
Static analysis
No suspicious patterns detected.