ClawHub
Back to skill

Security audit

AI Task Router Auto-Route Prompts by Complexity (Local/Cloud)

Security checks across malware telemetry and agentic risk

Overview

This appears to be a routing-focused skill with a scoping quality concern, but no evidence of hidden execution, credential theft, persistence, exfiltration, or destructive behavior.

Install only if you are comfortable with the skill influencing which model handles a request. Avoid using vague routing rules for confidential, regulated, credential-bearing, or high-risk work unless you have explicit model-selection safeguards and know where prompts are being sent.

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
74% confidence
Finding
The routing criteria are underspecified and based on broad examples, which can cause sensitive or complex prompts to be misclassified to weaker or local models. In a task-routing skill, ambiguous boundaries are security-relevant because they can lead to unintended handling of confidential data, poor-quality responses for high-risk tasks, or policy bypass through prompt phrasing.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.