ClawHub
Back to skill

Security audit

Workspace Guardian

Security checks across malware telemetry and agentic risk

Overview

The skill appears to provide workspace-organization guidance with disclosed file-management behavior, but users should be aware that its activation phrases are broad.

Install this only if you want an agent to help govern or clean up a project workspace. Before allowing it to move files, run cleanup scripts, or stop running tasks, ask the agent to show the exact planned actions and keep the scope limited to the current project.

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 (2)

Vague Triggers

Medium
Confidence
88% confidence
Finding
The trigger list is broad and maps to very common user requests such as 'clean up', 'project setup', and 'where should I save', which can cause the skill to activate in contexts the user did not specifically intend. Because this skill includes guidance to move files and run cleanup/validation scripts, unintended invocation could lead to unexpected file operations or policy enforcement on unrelated tasks.

Vague Triggers

Medium
Confidence
96% confidence
Finding
Using the plain word `stop` as a global emergency trigger is overly broad and can be activated accidentally during ordinary conversation, quoted text, logs, or task content. This can cause unintended termination of commands, sub-agents, or workflows, creating a denial-of-service condition and making the agent easy to disrupt.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.