Attention protected
One. Then done.
You always know your next move. Do it. The next one is ready.
01 · The problem
Too much at once. No clear next thing.
Work arrives faster than it gets finished. It lives in chat scrollback, in notes, in someone’s head. Every tool wants your attention, and the simple question - what should I do right now? - has no honest answer.
AI was meant to help. Mostly it adds another voice to the noise.
02 · What 123.do does
One next thing. Then the next.
Your work lives in Tracks. Each Step in a Track is in one of three places.
One Step in NOW per person. Everything else waits in DO, in order. Finish the one thing and the next is ready. No streaks, no scores, no guilt - just the next move.
03 · The team
Agents do the boring parts. You decide.
People and AI agents work on the same Track, under the same rule. Mention @b1ll on a Step and he does the work in his own NOW, then brings the result back to you with the next move already laid out.
d1on finds the right agent for each job so you don’t have to. m1ra sums up what they did in two sentences, so you read the state, not the feed. Pause d1on on any Track and nothing starts on its own.
04 · Your models
Bring your own AI. Keep your history.
Connect OpenAI, Anthropic, or any compatible provider on your own keys. You pay them directly, at cost. You bring the model; 123.do brings the consistent agent layer that drives it.
That layer is your organisation’s memory, and it lives above the model: you are never locked into one LLM, and a cheaper model is one swap away. Your history is yours, in one place.
05 · Everywhere
Phone, terminal, web.
Pick up the same Track on any device. Same NOW, wherever you look.
Keyboard-first CLI and TUI. Stays in your editor.
Watch the team move Steps as it happens. No refresh.
Focus on one thing. Finish it. Then the next.
A calmer way to work with people and AI.