Point an agent at the sources you follow, tell it what you care about, and a briefing arrives each morning. It is well written. It has headings. Anyone can have this now.
The morning, today.
The briefing has thirty items. Each is a fair summary of something that happened. Four of them matter to you, and you do not know which four until you have read thirty.
So you skim. A week later you stop opening it.
You try to fix it by talking to the agent. “Less funding news.” The next day there is none at all, and you miss the one round that mattered. Or nothing changes, because the chat you said it in is gone.
The summary was never the problem. The agent does not know which items you would have read, and you have no cheap way to tell it.
One list.
In Cube the briefing is a task that runs every morning. The agent keeps each story it considered as a row in a sheet: the source, the topic, one line on why it might matter to you. Then it writes the briefing from the rows it picked.
The run is one line in your list, beside everything else that is waiting for you. Open it and there is one page.
On top is the briefing. Below it, Cube lists anything the run wants your signature on, such as a message that passes one story on to someone you work with. Many mornings there is nothing to sign. Then the two buttons on the line mean something simpler: this was useful, or it was not. Cube keeps both answers.
Talk and sign.
You do not have to read thirty items to judge them. Talk to the page.
“Make this a list I can tick.”
It becomes one, right there. You tick four.
“Keep the ones I ticked, drop the rest.”
The page is four items long.
“Tell me more about the second one.”
The agent expands it in place.
“From now on, no funding rounds, except companies on my list.”
That sentence is about every morning after this one.
What freezes.
A chat would say “got it”, and then forget or overdo it. Cube writes your sentence as a change to the task and puts it under the reply for you to sign.
Before you sign, you see the rule in your own words and what it would change. If you have judged stories of that kind before, you also see how the rule would have handled them, each beside what you decided. If it would have dropped something you kept, Cube does not propose it. It tells you that your words and your choices disagree.
Sign it, and from the next morning those stories are not in your briefing. The page carries no note about what was left out. The count is on the task’s page.
The shape of the page can settle too. If you keep asking for the same layout, Cube can propose to show you this briefing that way from now on. That is a change to the task as well, and you sign it or skip it.
A rule you signed runs as a program: the same way every morning, for almost nothing. What it is not sure about still goes to the agent. You can roll it back.
What stays yours.
The story no rule fits. A name you have never seen, in a field you started watching last week. The exception you made on purpose, like the one funding round you did want.
And the question underneath: what you are watching for at all. A rule covers what you have already seen. What deserves your attention next is still yours to say, in one sentence, when it changes.
What it costs you.
I am not going to print a number here. I do not have yours.
The first mornings cost the most. You read the whole briefing, tick what you want, and say a sentence or two about the rest. Each sentence you sign shortens every morning after it.
The task’s page keeps the count: how many stories the rules left out on each run, and what share of them a program now handles. A briefing is working when you read all of it and still have most of your morning.
More use casesLead generationJob search