需求管理

Security checks across malware telemetry and agentic risk

Overview

This demand-management skill is coherent and purpose-aligned, with the main privacy consideration being that approved demand details can be shared with other workflow skills.

Install this only in an environment where downstream workflow skills and event listeners are trusted, because demand titles, descriptions, review comments, and prompt templates may be included in approval events and logs.

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 (1)

Description-Behavior Mismatch

Medium
Confidence
84% confidence
Finding
The skill emits a cross-skill event with the full `demand` object, which may include free-form business descriptions, reviewer comments, and prompt templates that downstream listeners do not need. In an event-driven skill ecosystem, broadcasting unnecessary sensitive context expands the trust boundary and can leak internal data to other skills or logs without explicit authorization.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal