ClawHub
Back to skill

Security audit

voice-review

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed writing-review helper, with some routing and auto-editing cautions but no evidence of hidden, destructive, or unrelated behavior.

Install this if you want an automated writing voice-review workflow and are comfortable with it reading the text you provide, applying limited hard-failure edits, and saving the final result. Prefer invoking it explicitly as /voice-review, review the diff after hard fixes, and enable learning mode only for text you are comfortable storing in local voice-profile snapshots.

SkillSpector

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 (3)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The trigger list contains generic terms like 'voice', 'review', 'prose', 'craft', and 'quality', which are broad enough to cause accidental invocation in unrelated contexts. In an agent ecosystem, unintended activation can expose user content to automatic processing, edits, or downstream agent dispatch without the user explicitly intending to run this skill.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill explicitly states that hard failures are 'auto-fixed' silently and later documents saving text and snapshots to disk, but it does not prominently require informed user consent before modification or persistence. This is risky because it can alter user-authored content and retain potentially sensitive text in files or learning snapshots without a clear warning at the point of action.

Shadow Command Trigger

Medium
Category
Trigger Abuse
Confidence
95% confidence
Finding
Using 'review' as a trigger conflicts with a common built-in command namespace and can cause shadowing or ambiguous routing. That increases the chance that users invoke this skill when they intended a trusted built-in review flow, leading to unexpected agent dispatch, auto-edits, or file persistence under this skill's logic.

Static analysis

No suspicious patterns detected.