Habits

Overview

By the time you sit down with your coffee, something has already read the night's mail. A habit is what your second brain does by itself: when something arrives, or at a moment you set, it reads what is there and works out what it might mean — and then stops, because everything it produces is a proposal .

What it applies is the same judgement these docs teach you. It asks which node the item hangs on , searches what you already keep before it proposes anything new, and leaves routine alone. A habit is your filing judgement running ahead of you — which is why teaching it is worth more than any other setting.

One comes with every account, and it is always on: the one that runs your inbox . It reads everything that arrives — mail, events, what other systems push in, the thoughts you put down — and works out one proposal per subject. Two mails and the meeting they are about are one matter, so they become one proposal; a bug report and a quote from the same client are two. It is not listed among your habits: what it does is written on the inbox itself, beside what it proposes.

It waits a moment before it looks, so what arrives together is read together. A reply in a thread it already has a proposal for goes straight to that proposal, so the proposal never lags behind the conversation it is about. Newsletters, receipts and other items that ask nothing of you get no proposal at all: they are marked as looked at and stay in your inbox.

Basics

Teach the inbox

The inbox keeps its instructions beside its proposals: one document, in plain sentences, that saves itself as you write and applies from the next item that arrives. Your second brain already knows the general rules — what earns a node, what hangs where — so what belongs here is what it cannot work out: your exceptions, and the way you do things. A heading per kind of item keeps it readable:

• Mail — "mail from the school is Family, and hangs on the school node", "never draft a reply to a newsletter".
• Calendar — "the Tuesday sync never needs a task", "interviews always get a prep block, due the day before".
• Signals — "alerts from staging need nothing", "a GitHub issue hangs on the project it names".

Several mailboxes, or several signal types that each want something else? Give each a heading of its own in the same document — "Mail on work@", "GitHub issues". It is the one place the inbox takes its instructions from.

What you let go of writes the instructions. A proposal you let go of for the same reason every week is one sentence here — that is the loop that makes month three quieter than week one. And keep the boundary with context : an instruction says what to do with what arrives; what is true about you everywhere belongs in the note every habit and every chat reads. When you send one of its proposals back with a remark that holds for good, your second brain proposes the sentence for these instructions itself, under the heading it belongs to, and it applies once you carry the proposal out.

Ask it to look again

You can ask for a fresh look by hand on any item, and it happens at once — an archive is a verdict on what you saw, not a refusal to look again. When an item landed in the wrong proposal, say so when you send the proposal back: the item leaves it and is looked at on its own.

Make a habit of your own

The inbox answers what arrives; a habit of your own answers a moment. Setting up the week every Monday morning, preparing what a node needs when its status moves to Review — work you would otherwise remember to start yourself. Your habits have a place of their own, beside the inbox: one list, by name, where a habit you switched off keeps its place, faded. Make a new one with the + on that list: a name, and its instructions if you have them already.

Every habit is then a page, the way a node is. Its instructions are the page itself — a document, written as long as it needs to be, that saves itself as you write. Beside it stand its properties, each one a change that applies at once:

• On or off — a habit that is off stays where it is, with everything you wrote on it.
• When it kicks in — on a schedule, like every Monday at eight, or when a node's status changes, with an answer of its own per status.
• How smart it is — how much thought it brings, from quick to thorough. The default suits most.
• What it may reach — which of your connected tools it may use. A new habit reaches none.

Under What it did stands every time it kicked in, so you can read back what it saw and what it proposed. Its actions, deleting it among them, close the column.

Know what it may touch

A habit carries scopes: what it may read, and what it may propose. Read access covers what it needs to judge with — your nodes, your mail, your calendar, your context note. There is no scope that lets it change something without asking, because a write is a proposal here.

The inbox's scopes are fixed to what reading an inbox needs, which is why what you give it is instructions and nothing else. A habit you make yourself is where you choose which tools it may reach.

FAQ

Can I switch the inbox off?

No. Reading what arrives is what the inbox is for, so it is always on. A habit you made yourself switches off and on from its page, and keeps its instructions while it is off.

Why did nothing happen for an item?

Proposing nothing is an answer. Routine rhythm — a daily standup, a block you made for yourself, a newsletter — is read and deliberately left alone.

Where do I see what it proposed?

Under Proposals , the first page of your Inbox , and on the page of every item a proposal is about. A proposal is a page, not a conversation: say under Anything else? what you would have different, or ask a chat to change it.