Async Digital / the spoken cue
How we work

The spoken cue

At a stopping point, or when a long task passes a milestone, a single line is spoken aloud that says where things are. It is written fresh each time to fit the moment, not a canned chime. It is a status signal you hear rather than read, so it reaches you across the room when you are not watching the terminal.

One line, spoken, bespoke to what just happened.

What it is

A line written for the moment

The cue is not a fixed sound. It is a short sentence, composed for the exact moment it marks, that names what just happened and what comes next. It is spoken at two kinds of point: at a stop, when a turn ends and attention is wanted or the work is done, and at a milestone, when a long run crosses something worth knowing about without waiting for the end.

MilestoneTests are green, opening the pull request now.
StopBlocked on a review verdict, so I have moved to the next task.
StopHandoff is written and the branch is pushed. That is a clean stopping point.

Because each line is written for its moment, it carries real information, not just the fact that something occurred. You learn what occurred and whether it needs you, in the time it takes to hear one sentence.


Why it exists

A signal that reaches you when you look away

Much of the work runs while nobody is watching the screen: a long build, a background wrap-up, a task left to run while attention is elsewhere. Written status only helps someone who is reading it. A spoken line does not need to be read. It reaches you across the room, mid-conversation, or while you are on something else entirely, and it tells you whether the work still needs you.

It stays off the working context

The cue rides a channel separate from the work. It is spoken, not written into the transcript the session is reasoning over, so it does not accumulate in the context the work draws on the way a running status report would. A cheap, out-of-band signal that costs the working context nothing.

It carries across sessions

Because it is ambient rather than tied to one terminal window, the cue reaches you the same way whichever session speaks it. Across many sessions running at once, the spoken line is a single, consistent channel: you hear the state of the work without hunting for which window it came from.


How it fits

The second channel

The terminal already carries a status line: a compact, glanceable marker of what the system is doing at each checkpoint. That is for the eye, when you are looking. The spoken cue is the second channel, for the ear, when you are not. They do not repeat each other; they cover different moments.

For the eye

The terminal status line

A compact marker at each checkpoint, read at a glance while you are watching the work happen.

For the ear

The spoken cue

A single spoken line at a stop or a milestone, heard when you are not watching, so nothing waits silently on you.

Together they mean the work can run unattended without going dark. When it needs you, it says so, out loud, in one line. When it does not, it tells you that too, and you can leave it be.


Go deeper