Loop
Security checks across malware telemetry and agentic risk
Overview
This instruction-only skill is a bounded retry helper that logs progress locally and does not contain hidden code, though users should review each task and command it repeats.
Install this if you want the agent to retry clearly scoped tasks until defined criteria pass. Keep success criteria narrow, review commands before they run, require explicit approval for destructive operations, and avoid putting secrets in loop notes because history is saved locally under ~/loop/.
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
62/62 vendors flagged this skill as clean.