intimate-ch06-interdependence
Security checks across malware telemetry and agentic risk
Overview
The supplied scan signals do not show malicious or suspicious behavior, but the actual target skill artifact was not clearly present in the workspace for independent confirmation.
Based on the supplied scan data, this skill does not need to be held for review. Because the target artifact was not clearly available for direct inspection, users should still review the skill text at install time and avoid granting credentials or broad local access unless the skill clearly explains why it needs them.
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.