This is going to sound very critical and it isn't meant to be - I absolutely love BTT. But it often does some strange things or won't do other actions (like recognizing images). I was having trouble getting an action sequence to work only to discover there were strange changes to what I had set up. Here's (in part) what I originally set up:
Here's what I found a few days later having made no changes:
I also often have to run a sequence several times before it will work from start to finish. For example I have BTT set up to fill in login names and passwords but the images are sometimes not recognized and have to be set up again in order to work or an image will be clicked on but nothing happens.
This is, again, not meant to criticize BTT but my hope is that it can somehow be made more stable. Running MAC OS 15.3.1 on a 2020 MacBook Pro 16" (Intel), mostly on Firefox v135, and BTT 5.199.
BTT doesn't change action sequences on it's own, this looks like a failed copy paste maybe? Or do you have the experimental Dropbox sync enabled? In that case it could be some sync merge issue.
Image recognition is always complicated. Always keep in mind that transparency or sizing of elements can have a big effect and try to play with the thresholds offered in BTT. Try to find a small as possible area that still has some significant features to distinguish it. In general the image recognition is very stable in current versions but it can be tricky to setup correctly in some cases.
For the other parts, BTT just sends the commands to the apps, it doesn't get feedback from the apps about what happens. There can always be subtle differences, processing delays etc.. Thus when trying to script with keyboard shortcuts and image recognition the delays you add are super important to get right. This is not something that can be fixed in BTT, but I know it can be very tricky to setup right. If you can use Apple Script that is usually much more robust, but of course is not possible in all cases.
Thank you for the suggestions. The screenshots were of actions I set up manually (no copy and paste done) and I don't have Dropbox. On a couple of those lines there is a "-" character which I never use. It's been a very strange phenomenon.
As for the image recognition I have found using a larger area with more detail does help some. The other strange thing is that sometimes the image is recognized and I'll have a left click action on that image but the next action step doesn't happen unless I do it manually. Then the sequence will continue. Is there a timing parameter that I'm not getting quite right? Seems like BTT goes faster than the computer or screen or website can process. In a few instances putting a delay BEFORE the "wait for image to appear" action seems to help but not consistently. Are there any general timing suggestions you can make? How do I get those delays right?
