Back to skill

Security audit

Docker Compose Deploy

Security checks across malware telemetry and agentic risk

Overview

The skill appears to be a documentation-only fraud-detection helper with no executable payload or scanner findings.

Review the SKILL.md instructions to confirm they match the fraud-detection workflow you expect, but the supplied scan context shows no executable payload, no VirusTotal signal, and no advisory findings requiring 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

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.