观势 — 数据分析专家

Security checks across malware telemetry and agentic risk

Overview

This is a coherent data-analysis skill with a disclosed dependency-installation caveat but no evidence of hidden, destructive, or exfiltrating behavior.

Before installing, treat the automatic dependency installation as the main operational risk: run it in an isolated environment, approve any package installation yourself where possible, and avoid using it on sensitive datasets unless you are comfortable with the agent reading those files and writing analysis outputs.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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)

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The skill instructs automatic `pip install` into the current Python environment when dependencies are missing. That expands the skill's behavior from data analysis into environment modification and arbitrary package retrieval, which can introduce supply-chain risk, break isolation, or alter other tasks sharing the same runtime. In this skill context, the behavior is not strictly necessary and is more dangerous because the toolchain includes code execution via `python_executor`.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal