I also use: qrun (for running notes, Standard Operating Procedures), qlog (for logs of activities on projects, journal entries), qtalk (for lectures), qbug (for bugs, things to fix or tweak, etc.),… You get the idea. Adapted this from Merlin Mann’s “runx” convention, but moved the odd character to the front to take advantage of autocompletion.)

Another Authotkey user here! 😃 On my machine, Mehul's solution with the 1ms delay takes noticeably longer to actually insert some text. I found the following solution which inserts it in a less "chunky" manner. Adjust the 1ms till it works for your setup :) ; This worked ::dx::{Sleep 1}DevExpress ; For longer hotstrings, I needed more ::azerty::{Sleep 60}DevExpress With 250ms pretty much any length of hotstring expanded correctly. I answered this Stackoverflow question with details from this issue. The same bug might have been present in an earlier version of VSCode: microsoft/vscode#1934 They make mention of this commit fixing it. Unfortunately it's a rather large commit :( microsoft/vscode@a1bd50f Commit msg: "Fixes #1168: Read synchronously from textarea" The problem has to do with the backspace remapping. Take the following autohotkey hotstring: ::tada::🎉 This will make typing "tada" followed by one of the "EndingChars" (space, tab, comma, dot, ...) expand to the 🎉 emoji. What you see visually happening on the screen is that Autohotkey does this by first sending a backspace to the editor 4 times (length of hotstring "tada") and then inserts the replacement text (🎉) What happens when this (pretty fantastic) extension is active is that the first x characters get deleted then the replacement text gets inserted and then the remaining (hotstring length - x) characters get deleted. But because the cursor is now at the end of the replacement text... which gets chewed on 😃 I'll have to learn how to debug the IDE itself or always add a {Sleep 250} to my hotstrings...