Task Router

Security checks across malware telemetry and agentic risk

Overview

This skill is a simple local task classifier that prints routing suggestions and does not itself run commands, trade, access credentials, or send data out.

Safe to inspect or use as a local suggestion tool. Before connecting its output to trading, system execution, cron, deployment, or other action-capable agents, add explicit confirmation, allowlists, and separate checks for sensitive or destructive requests.

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 skill advertises automatic routing for a wide range of tasks using broad intent categories and keyword-based matching, but it does not define clear activation boundaries, authorization checks, or safety gating before dispatch. In an agentic environment, this can cause sensitive, risky, or out-of-scope requests to be forwarded to powerful downstream tools or specialized workflows without sufficient validation, increasing the chance of privilege misuse or unsafe tool invocation.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal