First, there are no APIs for double-tap. Anything double-tap does Apple implements, developers don’t have anyway to assign anything to it.
Second, even if there was a way for developers to use double-tap, it would not be applicable here because dictation on watchOS is a system-level function. Apps just tell watchOS they would like text input, and watchOS pops up the input sheet run by watchOS, in a totally separate process from the app, and then hands back the text to the app after the “done” button pressed, so there would be no way for me to intervene there.
Hopefully, that is something Apple decides is worth changing.
This is great! Thank you. For me, when I run it using the action button on the Ultra 2, after each recording, I am prompted with ‘Allow “Dictate 2 Drafts” to copy to the clipboard?’, then ‘Allow “Dictate 2 Drafts” to share 1 Speech Recognition with “Drafts”?’ Even when I tap “Always Allow” on the second question, I still get both prompts every time. I reset the privacy settings for the shortcut on the iPhone, but that didn’t change anything. Is there a fix for this?
While I’m asking questions, can a step be added to the shortcut that takes it back to the Home Screen after completing the shortcut? I am left on the Shortcuts screen after completion.
I had this problem as well. kept me from using it. I think I just solved it. I ran the shortcut ONCE on my iphone and did the Always Allow there for the 2 prompts that showed every time (clipboard and paste dictation text). Now it works on watch without prompting. It seems like the decisions made from watch side aren’t “sticky” but they are on the iphone.
I must thank you for this tip. I run beta and every time a new beta would come I would check to see if clicking always allow on the watch would stick and it never did. Came back here to see if anyone figured out a solution. Glad I did. Thanks! My action button is now useful.
I think it’s a Shortcuts bug with values getting passed around when triggered from the Watch. I think you can also workaround it by using a variable step in-between such that the result of the dictate gets set in a variable, then the variable value used to create the draft.
That said, I think the below tip is a better solution for doing capture via the action button on a Ultra now that watchOS 11 remembers that you prefer dictation for text input and you don’t have to hit the mic button every time (and it runs all on the watch and doesn’t need the phone):
