What part are you wondering about exactly?
If it were me, I'd start by breaking down the tasks. An easy part would be the basic UI setup. Just some control nodes and options to allow typing and show output somehow.
Another part would be algorithms/manipulation to get the elements to autocomplete to. What is the set of commands the user can do that they should see?
By this time, updating the UI to reflect those completions will probably be pretty clear. How can you show the autocomplete options in an intuitive way and allow the user to choose it?
I bet there are examples of basic auto completion on the internet in lots of languages, especially a shell language like Bash. After that, translating to GDScript shouldn't be too bad