竞品情报监控系统

Security checks across malware telemetry and agentic risk

Overview

This skill is unlikely to harm your computer, but it appears to sell competitor research while generating mostly generic and partly random reports.

Treat this as a Review item before installing. It does not show malware-like behavior, but do not rely on its reports for pricing, marketing, investment, or product decisions unless the publisher clearly labels the output as synthetic/template-based or adds real data collection, source attribution, and deterministic review analysis.

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
96% confidence
Finding
The skill advertises competitor analysis from a provided name/link, but the implementation never uses the URL to retrieve or validate real competitor data. This can mislead users into trusting fabricated output as evidence-based intelligence, creating integrity and decision-risk issues in a business-analysis context.

Intent-Code Divergence

Medium
Confidence
98% confidence
Finding
The review-analysis function always generates random sentiment results via Math.random() and ignores any actual review input, yet presents the output as analysis. In this skill’s competitor-intelligence setting, that is dangerous because it produces nondeterministic, invented business signals that may drive pricing, marketing, or product decisions.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal