Press ⌘K, or Ctrl+K away from a Mac, and a box opens over whatever you were doing.
That box is the fastest route to most things in NearSync, and the reason is not that it searches well. It is that it does three different jobs without asking you which one you want.
Three lanes, one box
Go. Find something and open it. A deal, a person, an invoice, a message, a screen.
Ask. Put a question to the assistant, against your live data, and read the answer inside the same panel.
Do. Say what you want to happen in plain words, and get a task, a meeting or a reminder out of it.
You do not choose a lane. You type, and the box works out from what you typed which lane leads. Typing a name searches. Typing a question offers to ask. Typing something with a time in it offers to create.
Why this matters more than it sounds
Every platform has a search box, and most of them are a worse version of the navigation menu.
The difference here is what the box can reach. It searches twelve different kinds of thing at once, across parts of the platform that in most stacks are separate products: your deals and your contacts, but also your tasks, your Cockpit queue, your chat messages, your invoices, your engineering issues, your knowledge base and the people you work with.
That is possible for the same reason the Cockpit is possible. It is one database, so one query reaches everything, and there is no integration to keep in sync and nothing to be out of date.
The classification is deliberate, and dumb
The decision about which lane leads is made by simple rules rather than by a model.
A question mark, or a sentence starting with who, what, why, how, when, where, which, can, does or should, leans towards asking.
A date, a time, a duration or a priority marker leans towards creating.
Anything else is neutral, and the search results lead.
Nothing is sent anywhere to make that decision, which is why the box never pauses to think and why typing costs nothing. The assistant is engaged only when you deliberately take the Ask row, and the credits are spent then rather than on every keystroke.
What is on screen before you type
The box is not empty. It opens onto a short deck: where you can jump to, what the assistant can be asked, what plain words can create, what is happening in the next few minutes, where you were last, and the conversations you have open.
That deck is the answer to "what is this box for", shown rather than explained, and it has its own piece.
What it will not do
It will not find things by description. Searching "overdue invoices" matches an invoice literally called that, not the set of invoices that are overdue. For a question like that, take the Ask row; it is the right tool and it is one row down.
It will not widen your access. Everything it reaches, you could already reach. Private channels you are not in, colleagues' tasks and records outside your permissions do not appear here, and there is no setting that changes it.
It will not return everything. Each kind of record contributes a handful of matches. It is built to get you to a known thing quickly, not to produce a list you can work through.
It will not remember by default. A question asked here is not saved unless you say so, which is a feature rather than an omission and is covered in the piece on asking.
Who gets the most out of it
Anybody who knows what they are looking for. That is the honest test. The box beats the menu when you can name the thing and loses to it when you are browsing.
Anybody who captures. The Do lane is worth more than the search to people whose problem is remembering rather than finding, and it is the half most people never discover.
Anybody on a laptop all day. The saving is a few seconds each time and thirty or forty times a day.
Where it fits against everything else
Against the sidebar: the sidebar is for browsing when you do not know what you want. The box is for when you do.
Against Conversations: the same assistant, in a panel rather than a page. Ask here for a quick answer you will not need again; go to Conversations for a working session you want kept.
Against the Cockpit: the Cockpit is what wants you. The box is what you want.
The first week
Day one, use it once instead of the menu. Pick something you know the name of. That single journey is the whole argument.
Day two, notice the deck. Open it with nothing typed and read what is there. Most people have no idea the box shows their next meeting until somebody points at it.
Day three, capture one thing with ⌘;. A task with a time in the sentence. It is the feature with the steepest drop in effort and the lowest discovery rate.
By the end of the week, the two chords are muscle memory and the menu is for browsing only.
The habit worth forming
Stop navigating. Almost every "open the thing I am thinking about" journey is faster from here than through the menu, and the difference compounds across a day.
The specific test: if you know the name of the thing, do not go looking for it. Two keys and four characters is quicker than any menu, and it works identically from every screen in the platform.
What to read next
Finding anything covers what the search actually reaches and how results are ordered, which is the piece most people get value from immediately.
Creating from plain words is the one people do not discover on their own, and it is the half of the box that saves the most time once you know it exists.
Did this answer your question?
No, ask a person