Correct setup for conversational coding using a phone?

How do I set up a coding workflow where I am doing it on the phone without looking at the screen, using just voice? Is this possible?

I know that when using the mobile app I can talk into the keyboard but I still have to tap send and the reply from the agent is typed out, not spoken.

Is it possible to do it as a conversation? I talk, agent understands when I’m done talking, does some work, prompts me with questions, I reply… etc.

Obviously this will not do for everything. But I would love to do it this way while I walk, and I don’t know how to set it up.

1 Like

voice input is leveraging your phone’s speech-to-text, so you still have to tap send, and replies come back as text. There isn’t a hands-free “continuous conversation” mode where the agent automatically detects when you’re done speaking.

1 Like

Thanks, so it’s not just me not being able to find it. It seems like something that would not be a huge effort to build, so I hope they do at some point.

We are looking into TTS for Agent output on Cursor Cloud but there is no timeline yet. Great idea though. i’d post that in Ideas > Feature Requests to see overall interest in that