Subagent Sheepdog
Security checks across malware telemetry and agentic risk
Overview
This is a text-only workflow skill that helps an agent verify delegated or long-running work before reporting it as running.
Safe to install as a workflow aid. Expect it to make the agent more cautious and explicit when launching subagents, browser automation, or long-running tasks; it should not run code, collect secrets, or perform actions on its own.
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
65/65 vendors flagged this skill as clean.