Model Healthcheck

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward model availability checker, with the main caveat that it can actively test multiple configured models and may consume API resources.

Install this only if you want an agent skill that can inspect configured model identifiers and run short test sessions against them. Consider asking explicitly for a single model or confirming all-model checks first if API cost, rate limits, or disclosure of model error details matters in your workspace.

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: 90% confidence
Finding: The trigger phrase "check models" is broad and can plausibly match ordinary user requests about viewing, comparing, or discussing models rather than explicitly requesting an active healthcheck. Because this skill performs concurrent tests across configured models, unintended activation could cause unnecessary API usage, noisy side effects, and accidental disclosure of model availability or error details.

VirusTotal

62/62 vendors flagged this skill as clean.

View on VirusTotal