Base Network OSINT
Security checks across malware telemetry and agentic risk
Overview
The available evidence does not show hidden, destructive, or purpose-mismatched behavior; the only negative signal is a Unicode control-character prompt-injection indicator without supporting malicious artifact evidence.
Review the rendered skill text for unusual hidden or bidirectional characters before installing, especially around instructions and commands. Based on the supplied evidence, there is no supported reason to hold this skill for Review.
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
49/49 vendors flagged this skill as clean.