windows-automation

Automate Windows desktop tasks by launching apps, capturing screenshots, and simulating mouse and keyboard actions via PowerShell and Python scripts.

Security Scan

The skill's files, instructions, and requested capabilities are coherent with a Windows desktop automation tool — it performs screenshots, process listing, and mouse/keyboard simulation via PowerShell/Python and does not request unrelated credentials or remote installs.

Purpose & Capability

Name/description match the included Python scripts (app launcher, screenshot, mouse, keyboard). The required actions (launch apps, capture screen, simulate input, list/kill processes) are appropriate for a Windows automation/RPA skill.

Instruction Scope

SKILL.md instructs the agent to run the local scripts and use PowerShell to list processes, capture windows, and save screenshots to the user's Desktop. Those actions are expected for automation but are sensitive (process enumeration, screenshots, simulated input). Instructions do not attempt to read unrelated config or secrets and do not transmit data externally.

Install Mechanism

No install spec or remote downloads are present; the skill is instruction-plus-local-scripts only. All code is included in the bundle, so there is no hidden remote install risk.

Credentials

The skill requests no environment variables or credentials. It performs filesystem writes (saving screenshots) and queries OS process/window state — these are proportional to its stated purpose.

Persistence & Privilege

The skill is not always-enabled and does not request elevated platform privileges in metadata. Some operations (killing processes, controlling other apps) may prompt for administrator rights on Windows; also, because the platform allows autonomous invocation by default, consider the risk that an agent could run these actions without explicit consent.

Assessment

This skill appears internally consistent for Windows desktop automation, but it can perform sensitive actions (take screenshots, enumerate processes, simulate mouse/keyboard, kill processes). Before installing: (1) review the included Python files yourself (they’re present and readable), (2) run the skill in a test or sandboxed Windows account/VM first, (3) avoid granting administrator privileges unless you trust the code, (4) restrict or review any agent autonomy so the agent cannot run desktop-control commands without your confirmation, and (5) prefer skills with clear source/homepage provenance — this package has unknown origin.

Windows Automation Skill

Overview

This skill enables Windows desktop automation through integrated PowerShell and Python scripts. It provides capabilities for launching applications, capturing screenshots, and simulating mouse/keyboard interactions. This skill is ideal for RPA (Robotic Process Automation) scenarios, automated testing, repetitive task automation, and controlling Windows applications programmatically.

Trigger Phrases and Execution Logic

Primary Trigger Phrases

Application Control:

• "打开 [应用名称]" (open [app name])
• "启动 [应用]" (launch [app])
• "关闭 [应用]" (close [app])
• "打开网站 [URL]" (open website [URL])

Screenshot:

• "截图" (screenshot)
• "截屏" (capture screen)
• "保存屏幕截图" (save screenshot)
• "截取窗口 [窗口标题]" (capture window [window title])

Mouse Operations:

• "移动鼠标到 (x, y)" (move mouse to)
• "点击" (click)
• "右键点击" (right click)
• "双击" (double click)
• "拖拽从 (x1, y1) 到 (x2, y2)" (drag from to)
• "滚动 [上/下]" (scroll [up/down])

Keyboard Operations:

• "输入文本 [文本]" (type text)
• "按下 [按键]" (press key)
• "组合键 [按键1] [按键2]" (combo key)
• "快捷键 [按键组合]" (hotkey)

Core Capabilities

1. Application Launching (app_launcher.py)

Execute application control operations through the app_launcher.py script.

Start an application:

python scripts/app_launcher.py start "C:\Program Files\Notepad++\notepad++.exe"

Open URL:

python scripts/app_launcher.py url "https://www.google.com"

List running processes:

python scripts/app_launcher.py list

Kill an application:

python scripts/app_launcher.py kill "notepad"

When to use:

• User wants to open a specific application
• User wants to launch a website
• User needs to manage running processes
• User wants to close/terminate applications

2. Screen Capture (screenshot.py)

Execute screenshot operations through the screenshot.py script.

Capture full screen:

python scripts/screenshot.py screen "C:\Users\YourName\Desktop\screenshot.png" 0

Capture specific window:

python scripts/screenshot.py window "Notepad" "C:\Users\YourName\Desktop\notepad.png"

List all visible windows:

python scripts/screenshot.py list

When to use:

• User wants to capture the entire screen
• User wants to screenshot a specific application window
• User needs to list all visible windows
• User wants to document UI states

3. Mouse Control (mouse_control.py)

Execute mouse operations through the mouse_control.py script.

Move mouse to position:

python scripts/mouse_control.py move 500 300

Click at position:

python scripts/mouse_control.py click left 500 300 1

Drag and drop:

python scripts/mouse_control.py drag 100 100 500 500 1.0

Scroll:

python scripts/mouse_control.py scroll down 5

Get current mouse position:

python scripts/mouse_control.py position

When to use:

• User wants to automate UI interactions
• User needs to navigate application interfaces
• User wants to perform drag-and-drop operations
• User needs to position mouse for other operations

4. Keyboard Control (keyboard_control.py)

Execute keyboard operations through the keyboard_control.py script.

Type text:

python scripts/keyboard_control.py type "Hello, World!"

Press single key:

python scripts/keyboard_control.py key enter

Press combination keys:

python scripts/keyboard_control.py combo ctrl c

Press hotkey:

python scripts/keyboard_control.py hotkey ctrl shift esc

When to use:

• User wants to automate text input
• User needs to simulate keyboard shortcuts
• User wants to fill forms automatically
• User needs to navigate applications via keyboard

Workflow Examples

Example 1: Automated Form Filling

User Request: "帮我打开Chrome,打开一个表单页面,填写姓名和邮箱"

Execution Logic:

1. Launch Chrome browser using app_launcher.py start chrome
2. Navigate to form URL using app_launcher.py url <form_url>
3. Wait for page to load
4. Use mouse_control.py move to position cursor in name field
5. Use keyboard_control.py type to enter name
6. Use mouse_control.py move to position cursor in email field
7. Use keyboard_control.py type to enter email

Example 2: Application Testing

User Request: "测试记事本应用,输入一些文字,截图保存"

Execution Logic:

1. Launch Notepad using app_launcher.py start notepad
2. Wait for application to be ready
3. Use keyboard_control.py type to input test text
4. Capture window screenshot using screenshot.py window "Notepad"
5. Report results and screenshot path to user

Example 3: Data Entry Automation

User Request: "自动打开Excel,输入一批数据到表格"

Execution Logic:

1. Launch Excel using app_launcher.py start excel
2. Open specific file using keyboard_control.py combo ctrl o
3. Use keyboard_control.py type to enter file path
4. Use keyboard_control.py key enter to confirm
5. Navigate to cells using mouse_control.py move or arrow keys
6. Enter data using keyboard_control.py type
7. Save using keyboard_control.py combo ctrl s

Integration Guidelines

Script Execution

1. Always check script output: Each script returns JSON with success, message, and optional data fields
2. Handle errors gracefully: Check success field before proceeding with dependent operations
3. Use appropriate delays: Add delays between operations to allow UI to respond
4. Coordinate with window states: Ensure windows are in the expected state before interacting

Error Handling

• If app_launcher.py fails to start an app, check if the path is correct
• If screenshot.py fails, verify window title or check permissions
• If mouse_control.py operations don't work, check screen resolution coordinates
• If keyboard_control.py input is incorrect, verify key names are supported

Coordinate System

• Mouse coordinates are absolute screen coordinates
• (0, 0) is the top-left corner of the primary monitor
• Use mouse_control.py position to find current coordinates
• For multi-monitor setups, specify monitor index in screenshot operations

Key Naming Conventions

Letter keys: a, b, c, ..., z Number keys: 0, 1, 2, ..., 9 Special keys: space, enter, tab, backspace, escape, delete, insert, home, end, pageup, pagedown Arrow keys: up, down, left, right Function keys: f1, f2, ..., f12 Modifier keys: ctrl, alt, shift, win

Resources

scripts/

This skill includes four executable Python scripts for Windows automation:

• app_launcher.py - Application launching and process management
• screenshot.py - Screen and window capture functionality
• mouse_control.py - Mouse movement, clicking, and scrolling
• keyboard_control.py - Keyboard input simulation and key combinations

Note: Scripts are executed without loading into context window for efficiency. Read them only when patching or debugging is needed.

references/

Delete this directory if not needed. Currently contains placeholder references.

assets/

Delete this directory if not needed. Currently contains placeholder assets.