量子计算思维模拟器

Security checks across malware telemetry and agentic risk

Overview

This appears to be a local educational quantum-simulation skill with possible accuracy issues, but no evidence of hidden access, persistence, credential use, exfiltration, or destructive behavior.

Install only if you are comfortable treating the simulator as educational or approximate until its gate math is reviewed. Do not rely on its CNOT, Toffoli, or entanglement results for coursework, research, or engineering decisions without independent validation.

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 (2)

Intent-Code Divergence

Medium

Confidence: 97% confidence
Finding: The CNOT implementation does not model joint two-qubit state evolution and instead mixes the target qubit using control probabilities. In a simulator advertised as demonstrating superposition and entanglement, this can mislead downstream users, educational consumers, or agent logic into believing entanglement occurred when the state representation cannot support it, producing incorrect results and unsafe trust in simulated outputs.

Intent-Code Divergence

Medium

Confidence: 96% confidence
Finding: The Toffoli docstring promises a controlled-controlled-NOT gate, but the code applies a probability-weighted partial flip based on independent control amplitudes. This semantic mismatch can corrupt computations, invalidate demonstrations, and cause users or dependent components to rely on mathematically false behavior in a quantum simulation context.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal