Isolated Multi-Agent Control Plane

• Outcome: Pass
• Latest audit: May 28, 2026 at 11:31 AM UTC
• Version:
  0.1.2

Overview

This skill is a local multi-agent project blueprint whose installer and utilities match its stated purpose and do not show hidden network, credential, or destructive behavior.

Install it only into a project directory you are comfortable modifying, since it creates local agent workspaces, task/mailbox files, logs, and shared-memory state. Review the role rules if you plan to let agents act autonomously, especially around write access and human approval for deployment actions.

SkillSpector

By NVIDIA

Vulnerability Patterns

• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (1)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 89% confidence
• Finding: The skill advertises and instructs use of shell commands plus Python utilities that can read and write files, but it does not declare any permissions. That mismatch is a real security issue because consumers and policy engines may underestimate the skill's operational capabilities, especially since it installs files into arbitrary paths and operates a control-plane with mailbox, logs, and shared-memory storage.

VirusTotal

63/63 vendors flagged this skill as clean.