You can automate keyboard shortcuts in Windows by using PowerShell with the .NET
SendKeys class, but you must bypass Session 0 isolation. The most reliable method is to create a scheduled task that runs under your interactive user account rather than the system account. This ensures the script executes in the active desktop session where keystrokes are delivered correctly to foreground applications.
Understanding Session 0 Isolation
By default, tasks created in Task Scheduler run in Session 0. This session is isolated from the user’s interactive sessions (Session 1 and above) for security reasons. If you attempt to send keys using
[System.Windows.Forms.SendKeys] from a Service or System account, the input goes nowhere because there is no graphical desktop to receive it. To fix this, your scheduled task must be configured to run whether the user is logged on or not, specifically selecting the "Run only when user is logged on" option. This forces the task into the current user’s session, allowing
SendKeys to interact with open windows. For more details on common scheduling pitfalls, see our guide on
Powershell Script Not Running in Task Scheduler: A Troubleshooting Guide.
Creating the PowerShell Script
The core of the automation is a simple PowerShell script that utilizes the
SendKeys class. You do not need external libraries if you load the necessary Windows Forms assembly. Create a file named
AutoShortcut.ps1 and add the following code. This example demonstrates sending
Ctrl+C to the active window, which copies selected text.
# Load the Windows Forms assembly
Add-Type -AssemblyName System.Windows.Forms
# Define the shortcut sequence
$shortcut = "^c"
# Send the keystrokes to the active application
[System.Windows.Forms.SendKeys]::SendWait($shortcut)
In this script, the caret symbol (
^) represents the Control key. Other modifiers include
+ for Shift and
% for Alt. You can combine these, such as
%{F4} to simulate Alt+F4. Always use
SendWait instead of
Send to ensure the keystrokes are processed before the script continues, preventing race conditions with slow applications.
Configuring Task Scheduler
Open Task Scheduler and select "Create Basic Task". Name it something descriptive like "Automate Keyboard Shortcut". When prompted for the trigger, choose "One time" or "Daily" depending on your needs. In the Action step, select "Start a program".
For the program/script field, enter
powershell.exe. In the Add arguments field, you must specify the execution policy and the path to your script. Use the following format to avoid permission errors:
-ExecutionPolicy Bypass -File "C:\Scripts\AutoShortcut.ps1"
Crucially, click "Finish" and then right-click the new task to select Properties. Go to the General tab. Change the security options to "Run only when user is logged on". If you leave this on "Run whether user is logged on or not", the task will still fail to send keys due to the Session 0 isolation issue mentioned earlier. Also, check "Run with highest privileges" if the target application requires administrative rights to receive input.
Handling Common Errors
Even with correct configuration, scripts may fail silently. If the task runs but nothing happens, verify that the target application is actually running and has focus.
SendKeys does not bring windows to the foreground; it only sends input to whatever window currently has the keyboard focus. If you need to launch an application first, you must add a
Start-Process command followed by a
Start-Sleep to allow time for the UI to render before sending keys.
Another common issue is incorrect syntax for special characters. In
SendKeys, curly braces
{} are used for special keys like F1, Enter, or Backspace. Parentheses
() are used for combinations. For example, to press Enter, use
{ENTER}. To press Ctrl+Enter, use
^%{ENTER} is incorrect; it should be
^{ENTER}. Consult Microsoft documentation for the full list of virtual key codes. If you encounter execution errors during setup, refer to
Task Scheduler: Executing PowerShell Script with Errors for debugging steps.
Scheduling Multiple Actions
If your automation requires multiple steps, such as opening a file and then typing into it, you can chain commands in your PowerShell script. Alternatively, you can configure Task Scheduler to perform multiple actions. However, chaining within PowerShell is generally more robust because you can control timing between actions using
Start-Sleep. For scenarios requiring complex sequencing, see
Running Multiple Actions in Windows Task Scheduler: Wait or Kick?.
Frequently Asked Questions
Why does my script work manually but fail in Task Scheduler?
This is almost always due to Session 0 isolation. Manual execution runs in your interactive session, while Task Scheduler defaults to the system session. Ensure you have selected "Run only when user is logged on" in the task properties.
Can I use SendKeys to type into a UAC prompt?
No. User Account Control prompts run in a secure desktop session separate from the standard user desktop.
SendKeys cannot reach these elevated prompts. You would need to use alternative methods like providing credentials via the task's "Run as" field or using command-line switches that accept passwords.
How do I handle special characters like @ or # in my shortcuts?
Special characters are sent directly without braces. For example, to type "@", just include "@" in the string. However, if you are combining them with modifiers, ensure the order is correct. The
SendKeys class interprets standard ASCII characters directly, so no escaping is needed for most symbols.
What if the target application is minimized?
SendKeys will not work if the application is minimized or hidden. You must either restore the window using
AppActivate or ensure the application is already visible. Restoring windows programmatically adds complexity and latency, so it is best to keep target apps open and visible when the task runs.
Is there a safer alternative to SendKeys for automation?
Yes, for web browsers or modern UIs, consider using browser-specific APIs or accessibility tools like UI Automation.
SendKeys is legacy technology and can be unreliable with DPI scaling or high-DPI monitors. For simple desktop apps, however, it remains the most straightforward built-in solution.