gstack Prototype

Security checks across malware telemetry and agentic risk

Overview

This skill is a small, disclosed prototyping guide that tells the agent to create clearly marked throwaway code, with no hidden scripts or malware signals found.

Install this if you want the agent to quickly create disposable prototypes. Be explicit when you want production-quality work instead, and review or delete any generated prototype files after use.

SkillSpector

By NVIDIA

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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)

Vague Triggers

Medium

Confidence: 86% confidence
Finding: The skill description uses broad, natural-language triggers such as 'prototype this', 'let me play with it', and 'try a few designs', which can match many ordinary requests outside a narrowly scoped prototyping context. In an agent-routing system, this ambiguity can cause the skill to activate in situations where throwaway-code behavior is inappropriate, increasing the chance of unsafe code generation, unnecessary file creation, or bypass of more specialized skills with stricter safeguards.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal