Hey, thanks for the feature request. The logic makes sense: attach an image → explain with voice → send. For UI and code tasks, that really is more convenient.
This has already been discussed here: Speech-to-Text Functionality Issues with Image Pasting in Agent Mode. I’d recommend checking it out and adding your thoughts or details there so we can keep the signal in one place.
I can’t share an ETA yet, but I’ll post an update if we have news on this.