Test Upload 2026
Security checks across malware telemetry and agentic risk
Overview
This is a documentation-only test skill for validating publishing, with no executable code or sensitive behavior found.
This appears safe to install as a test or demonstration skill. It does not add useful agent capability, so install it only if you specifically want a ClawHub publishing-workflow validation artifact.
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
64/64 vendors flagged this skill as clean.