Tenqua OpticalQuantumSkill

Security checks across malware telemetry and agentic risk

Overview

This skill is a small local numerical simulator with no evidence of hidden access, persistence, data theft, or destructive behavior.

Install or run this only as a local demo. Review the Python script first, use a trusted Python environment with NumPy installed, and treat the reported kernel value as an approximate stochastic simulation result rather than a dependable scientific or production ML metric.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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
Findings (1)

Intent-Code Divergence

Medium
Confidence
92% confidence
Finding
The function advertises a quantum-kernel-style overlap computation, but the implementation actually produces a heuristic score derived from intensity differences after introducing random fiber phase drift. This mismatch can mislead downstream users into trusting the output as a principled kernel value, causing incorrect scientific conclusions, faulty model behavior, or insecure reliance in higher-stakes decision pipelines.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal