MCP协议配置

• Outcome: Pass
• Latest audit: May 28, 2026 at 7:29 AM UTC
• Version:
  1.0.0

Overview

This appears to be a disclosed MCP guidance skill with some powerful integration examples, but no evidence of hidden, destructive, or deceptive behavior.

Install only if you are comfortable reviewing the MCP servers you enable. Prefer least-privilege filesystem paths, narrow GitHub/API token scopes, and consider installing uv through a package manager or by inspecting the installer instead of piping remote code directly into a shell.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

• Confidence: 86% confidence
• Finding: The skill explicitly recommends network-connected and file-system-capable MCP servers such as web search and filesystem tools, but provides no warning about data exposure, permission scope, or the risk of unintended system changes. In an agent-skill context, this omission can lead users to connect powerful external tools without understanding privacy, exfiltration, or host-impact risks.

VirusTotal

65/65 vendors flagged this skill as clean.