Security audit

Control

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent desktop automation guide, but it needs Review because it encourages powerful screen, keyboard, mouse, and clipboard control with weak safety scoping.

Install only if you intentionally want an agent to control your desktop. Keep failsafe and approval mode enabled, avoid running with administrator privileges unless narrowly necessary, do not automate real passwords or secrets, and treat screenshots and clipboard reads as sensitive data that may expose private information.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Intent-Code Divergence

Medium

Confidence: 93% confidence
Finding: The skill claims that risky code was removed and safety was enhanced, yet the documentation still recommends unsafe operational practices such as disabling failsafe protections and running with elevated privileges. In a desktop automation skill with mouse, keyboard, screenshot, and clipboard control, these instructions materially increase the chance of harmful or irreversible actions and undermine the stated safety posture.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The trigger keywords are generic terms like 'mouse', 'automation', 'advanced', 'control', and 'desktop', which are likely to appear in ordinary conversation. This raises the risk of accidental invocation of a powerful automation skill capable of input simulation and screen access, potentially causing unintended actions on the host system.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The documentation advertises screenshot and clipboard functionality without a clear privacy warning, even though both can expose secrets, personal data, tokens, or confidential application content. In a desktop-control context, these capabilities are especially sensitive because they can capture whatever is currently visible or stored system-wide, often beyond the user's immediate awareness.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The examples demonstrate automated entry of credential-like data and saving screenshots to disk without explicit warnings about sensitivity, permanence, or unintended disclosure. Such examples normalize risky behavior and can lead users to automate passwords, capture sensitive screens, or persist confidential information in files without safeguards.

Static analysis

No suspicious patterns detected.