Agent Synthesizer
Security checks across malware telemetry and agentic risk
Overview
This skill is not malware, but it tells the agent to obey and run commands from a changing GitHub README without clear review or approval boundaries.
Install only if you are comfortable having the agent consult that repository, and require it to show you the exact README commands before running anything. Use a disposable or low-privilege environment for setup, and do not provide tokens or credentials unless the README need is clear and reviewed.
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.