ClawHub

Auto Monitor

Security checks across malware telemetry and agentic risk

Overview

This is a small monitoring skill, but it asks the agent to run ongoing checks and immediately auto-fix issues without clear scope, approval, or stop controls.

Review before installing. Use only in an environment where you intentionally want proactive system monitoring, and configure explicit targets, check frequency, alert destination, stop conditions, and read-only defaults. Require confirmation before any restart, deletion, configuration change, or other repair action.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (3)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The skill explicitly instructs proactive monitoring and reporting without a clear trigger boundary, approval model, or scope restriction. Broad autonomous activation can cause the agent to act outside user expectations, access system state continuously, and generate unsolicited actions or disclosures, which increases the risk of privacy, safety, and authorization violations.

Missing User Warnings

High
Confidence
98% confidence
Finding
The statement that issues can be 'automatically repaired immediately' authorizes remediation without warning, confirmation, or change-control safeguards. This is dangerous because autonomous fixes can restart services, alter configurations, or delete data in ways that disrupt production systems or exceed the agent's authorization.

Vague Triggers

Medium
Confidence
90% confidence
Finding
The description says the skill will proactively monitor system status and report without waiting for user instructions, but it does not define triggers, scope, frequency, permissions, or boundaries. That ambiguity can enable unintended autonomous behavior, over-collection of system data, or surprise actions that violate least-privilege and user-consent expectations.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal