[feature request] Add a mode to use a default query instead of speaking

[josuah]

For instance, "give me more detail about what is that image".

Then gestures like one tap on Frame would run that default query, saving from speaking out loud, which might be inappropriate in a quiet place. This is also faster to transfer no audio at all, and immediately start transferring the image.

Some extra gestures could further allow selecting among several default queries.

https://discord.com/channels/963222352534048818/984966420603482182/1208934320077799484 https://discord.com/channels/963222352534048818/984966420603482182/1225391329161383936