Agent Spawner

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward multi-agent task planner with no evidence of hidden data access, persistence, or exfiltration.

Install only if you want an agent to coordinate parallel sub-agents. When using it, give each sub-agent clear, non-overlapping file scopes, use timeouts, and review collected outputs before accepting changes.

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

Low

Confidence: 80% confidence
Finding: This markdown file contains activation-like guidance such as 'Simple subtask', 'Medium subtask', 'Complex subtask', and rules like 'Research subtasks always use cheapest model' without defining boundaries or negative examples. Those broad categories are ambiguous enough that common tasks could match multiple buckets, increasing the risk of inconsistent or unintended use of the guidance.

VirusTotal

67/67 vendors flagged this skill as clean.

View on VirusTotal