yolo-expert

Security checks across malware telemetry and agentic risk

Overview

This is a YOLO/Ultralytics help skill with disclosed source-code lookup and example commands, and no evidence of hidden execution or data theft.

Install this if you want a YOLO/Ultralytics reference assistant. Before letting an agent run commands from it, confirm whether it may clone or update /root/ultralytics, run compute-heavy training/export jobs, overwrite experiment folders, or save labels, crops, and model artifacts.

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: 89% confidence
Finding: The trigger condition activates on any YOLO/Ultralytics-related question, which is overly broad for a skill that also instructs the agent to consult local source code and memory before answering. In a multi-skill agent, this can cause unintended routing, unnecessary access to local resources, and interference with more appropriate domain-specific skills, increasing the chance of incorrect or over-privileged behavior.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal