Windows UI Automation

Automate Windows GUI interactions (mouse, keyboard, windows) using PowerShell. Use when the user needs to simulate user input on the desktop, such as moving the cursor, clicking buttons, typing text in non-web apps, or managing window states.

Security Scan

The skill's scripts and instructions match its stated purpose (Windows desktop automation via PowerShell); there are no hidden network calls or credential requests, but it can simulate user input so install only if you trust it and run on a Windows system (the metadata lacks an explicit Windows-only restriction).

Purpose & Capability

The name, description, SKILL.md, and included PowerShell scripts (keyboard_control.ps1.txt and mouse_control.ps1.txt) are consistent: they implement mouse movement/clicks and SendKeys for keyboard input and use user32.dll and System.Windows.Forms as expected. Minor inconsistencies: metadata lists no OS restriction despite the skill clearly targeting Windows APIs, and the scripts are shipped with a .txt extension (powershell -File will still run them but this is unusual).

Instruction Scope

Instructions are concrete and constrained to running the included PowerShell scripts and a simple AppActivate call. They do not reference external endpoints, credentials, or arbitrary filesystem reads. However, the capability itself (simulating mouse/keyboard and focusing windows) inherently allows an agent to drive the desktop and perform potentially sensitive actions (e.g., interact with other apps, paste text). This is expected for the stated purpose but increases risk if invoked without user oversight.

Install Mechanism

Instruction-only skill with no install spec and no downloads — lowest-risk installation surface. All code is provided in the skill bundle; no external packages or URLs are fetched.

Credentials

The skill requests no environment variables, credentials, or config paths. The scripts do not attempt to read other environment variables or secrets. The scope of access requested is proportional to desktop automation functionality.

Persistence & Privilege

always:false (default) and model invocation is allowed (default). Autonomous invocation is platform-default and not inherently suspicious, but because the skill can fully control the desktop, autonomous invocation increases the potential impact. There is no mechanism in the skill to persist or modify other skills' configs.

Assessment

This skill appears to do what it says: run PowerShell to move the mouse, click, and send keystrokes. Before installing: (1) Only use it on a Windows machine and preferably in a controlled environment (VM/sandbox) — the metadata does not declare an OS restriction. (2) Be aware it can automate any UI action your account can perform (don’t grant elevated privileges while it runs). (3) If you need to limit risk, disable autonomous invocation or require user confirmation before the skill runs. (4) Review the included PS1 code (it’s short and readable) and, if desired, rename the files to .ps1 and set execution policy / permissions to match your security posture. (5) If you don’t fully trust the skill owner, avoid installing or run it only in an isolated environment.

Control the Windows desktop environment programmatically.

Core Capabilities

• Mouse: Move, click (left/right/double), drag.
• Keyboard: Send text, press special keys (Enter, Tab, Alt, etc.).
• Windows: Find, focus, minimize/maximize, and screenshot windows.

Usage Guide

Mouse Control

Use the provided PowerShell script mouse_control.ps1.txt:

# Move to X, Y
powershell -File skills/windows-ui-automation/mouse_control.ps1.txt -Action move -X 500 -Y 500

# Click at current position
powershell -File skills/windows-ui-automation/mouse_control.ps1.txt -Action click

# Right click
powershell -File skills/windows-ui-automation/mouse_control.ps1.txt -Action rightclick

Keyboard Control

Use keyboard_control.ps1.txt:

# Type text
powershell -File skills/windows-ui-automation/keyboard_control.ps1.txt -Text "Hello World"

# Press Enter
powershell -File skills/windows-ui-automation/keyboard_control.ps1.txt -Key "{ENTER}"

Window Management

To focus a window by title:

$wshell = New-Object -ComObject WScript.Shell; $wshell.AppActivate("Notepad")

Best Practices

1. Safety: Always move the mouse slowly or include delays between actions.
2. Verification: Take a screenshot before and after complex UI actions to verify state.
3. Coordinates: Remember that coordinates (0,0) are at the top-left of the primary monitor.