Security audit
Agentnomos Governance Preflight
Security checks across malware telemetry and agentic risk
Overview
This is a read-only governance checklist skill that tells an agent when to allow, pause, or block risky actions without performing those actions itself.
This skill is appropriate if you want an agent to pause before high-impact actions and produce a governance checklist. It may cause the agent to hold or block actions involving money, production systems, sensitive data, deletion, signing, permissions, or outbound communication unless explicit evidence and approval are available. Treat its output as advisory, not legal, compliance, or production authorization.
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.