Huang Renxun Ai Strategy

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed, text-only AI/NVIDIA strategy framework with some finance-style analysis templates, but no code, tools, persistence, credential access, or hidden execution.

Install only if you want an educational NVIDIA/Jensen Huang AI strategy and market-analysis framework. Do not rely on it for investment decisions, and verify any current market data, quotes, or company claims independently before using its outputs.

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)

Description-Behavior Mismatch

Medium
Confidence
95% confidence
Finding
The skill is advertised as a narrowly scoped Jensen Huang/NVIDIA strategy advisor, but the appendix introduces reusable generic market and investment analysis report templates. Even with repeated disclaimers, this broadens the effective capability into quasi-investment-advisory content and can cause the agent to produce domain outputs outside its declared purpose and safety boundary.

Intent-Code Divergence

Medium
Confidence
93% confidence
Finding
The trigger section claims the skill should activate only for Huang Renxun/NVIDIA-related topics, but later content provides generic AI market-analysis structures that can answer broader investment-style prompts. This mismatch weakens containment and increases the chance of scope drift, where the agent follows latent templates rather than the stated activation boundary.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal