To prompt the user for input which can then be used by the script, you would need to customise the script itself to prompt for the value.
So, as an example, the following is a script called test_read.sh on a OneClick server running on Linux:
echo "Enter word:"
echo "You entered the word $DATA"
echo "Press the <Enter> key to continue:"
When run, it prompts the user to enter a word. The user enters a word and presses enter. It returns the word entered by the user and then waits for the user to press the Enter key to continue.
To use this, you would then create a custom menu item under the Tools menu. For example;
<item name="Test Script">
<command>/usr/bin/xterm -e /home/myuser/test_read.sh</command>
The example above requires that the user have xterm software running on their OneClick client system (i.e their PC). However, this can be changed to reflect whatever terminal software you use that has similar capability.
When run, you would see xterm appear as per the attached screenshots.
Enter a word and press enter, it will output that word. Then press Enter again and the xterm disappears.
This assumes the Linux DISPLAY environment variable is set to the users client machine (for example; export DISPLAY=localhost:0.0. You would need to set this in the .bashrc / .bash_profile or other environment settings file for whichever particular shell is being used on the Linux based OneClick server).
Another option is to write the script according to the OS and install the script on the client machine and use the <launch-application> tag. If the client OS is Windows, something similar to the following example could be used where <SCRIPT> would be the name of the script to run:
<command>cmd.exe /c start "My Script" cmd /c "<SCRIPT> && pause"</command>