ClawHub
Back to skill

Security audit

Custom Commands

Security checks for vulnerabilities and agentic risk

Overview

This instruction-only skill is not deceptive, but its broad backup, sync, delete, cloud, voice, and auto-suggestion behavior needs Review because safety boundaries are underdefined.

Install only if you are prepared to enforce safeguards yourself. Before any backup, sync, or clean action, require exact source and destination paths, choose the cloud destination explicitly, preview deletions or overwrites, and confirm what the skill may store in memory.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Confidence
91% confidence
Finding
The skill says to auto-suggest commands in relevant contexts, but it provides no eligibility rules, confirmation requirements, or boundaries on when suggestions should appear. In a command skill that includes backup, sync, clean, and audit actions, broad triggering can nudge users into unintended actions or cause command invocation from ambiguous context.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The voice command examples use natural phrases like 'backup my documents' and 'Write a 500-word article on AI ethics,' which are close to ordinary speech and may be triggered accidentally or via ambient audio. Because the skill exposes actions that can transfer or delete data, overly broad spoken activation increases the risk of unintended command execution.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The skill defines data-transfer and destructive commands such as backup, sync, and clean without warning text, scope limits, dry-run behavior, or confirmation requirements. If implemented as described, users may overwrite, exfiltrate, or delete files unintentionally, especially since backup defaults to cloud storage and clean operates on patterns that could match broadly.

Static analysis

No suspicious patterns detected.