User Permissions
Security checks across malware telemetry and agentic risk
Overview
This is an instruction-only safety skill that tells agents to ask for user approval before impactful actions and contains no executable code.
Installing this skill should make an agent more cautious about system-changing actions. For high-impact operations, ask the agent to restate the exact action and require a clear, action-specific approval before proceeding.
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.