Security audit
Multi Agent Collaboration
Security checks across malware telemetry and agentic risk
Overview
This is a text-only workflow skill for supervised multi-agent planning and review, with no evidence of hidden execution or data access.
Install this if you want structured human-in-the-loop task planning for AI agents. Keep task scopes explicit, review agent outputs before accepting changes, and avoid putting secrets or private internal details into reusable public templates.
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
VirusTotal engine telemetry is currently stale for this artifact.
Static analysis
No suspicious patterns detected.