The AI in your LifeOps

What it doesFive jobs with switches, and Chat without one

Settings → AI lists five jobs, one switch each, and each row names what it does and what it sends when it runs.

Life area filing is one job for three kinds of thing. A journal entry, a task or a note that has no life area is filed under one of yours, reading the thing's words and your areas with their descriptions, alongside a few of your own past filings and corrections as examples. Three sweeps do the work on their own clocks: entries every two minutes, taking only what you wrote in the last two days; tasks every two minutes, taking every open task with no area however old; notes every five minutes, taking every note with no area. Each thing gets one verdict, and a verdict of “none of these” is an answer about the areas that existed when it was given, so adding, renaming or removing a life area asks about unfiled tasks and notes again. An entry it could not place stays unfiled for good. Filing a thing yourself replaces the app’s pick, and that correction goes into the next ask as an example. A task under a goal skips the model and takes its goal’s area, which is why that half keeps working with AI switched off.

Area descriptions drafts what a life area you have only named means, written from the name and the rest of your list, for you to edit or clear. The sweep runs every two minutes and asks about each bare name once. On Settings → Areas a drafted description shows the word draft after it until you edit it, and editing it makes the words yours. An area you described yourself is left alone, and an area whose draft you cleared is not drafted again.

Transcribe a recording turns a video or a voice memo you post into a cleaned-up read-through of what you said. It runs off the recording as soon as it is posted, with no sweep, and needs two keys stored in Settings → Integrations, ElevenLabs for the transcript and Anthropic for the clean-up. It is the one job that sends a file rather than text. The row on the AI page is its only switch. See Your videos and photos.

What changed in a version writes, when a note, a goal or an event is sealed as a new version, the line saying what moved between it and the one before, and for a goal, whether the target itself moved and from what to what. It runs on the hourly pass that seals versions.

Quick Capture, Journal or Task decision decides, when you use Quick Capture on the Apps page, whether what you captured is a task or a journal entry, and trims a task to its own words with the lead-in dropped. It runs the moment you capture, with no sweep. A spoken capture is transcribed first and then decided the same way. With this job off, every capture is a journal entry.

Every job on that screen also names the model it runs on and the prompt it runs with, and both are yours to change.

The Model row on each job offers Opus and Haiku, and Opus is the default.

Prompt beside it opens the words the job runs with, which start as the ones this repository ships and are the whole of what is sent as instructions. Your edit replaces them until you clear it, and clearing it puts the shipped prompt back.

Chat is the sixth thing that runs on your Anthropic key, and it is not on that screen. The speech bubble in the top corner opens one conversation. Asking a question sends the recent thread, up to the last twenty-four messages, to Anthropic, along with a token that lets the model call back into your own backend. It can WRITE a task, a goal, a note, a journal entry and a record of the conversation, and it can change a task, a goal or a note it names by its id. It can only READ your day and your areas. It can also search the web and fetch pages, through tools that run on Anthropic’s side. Chat has no switch in Settings → AI or anywhere else: the master switch does not stop it and neither does any per-job switch, and removing your Anthropic key in Settings → Integrations is the only thing that does, after which it refuses instead of answering. It runs on claude-opus-5, the same Opus the rows above default to, set on the backend by LIFEOPS_ASSISTANT_MODEL, and the words it is sent with are written into the code. You cannot change either from the app. The user’s guide’s Chat page covers using it.

One thing the switches do not govern. Entries tagged Agent in your journal come from a different actor, your Claude, writing on your behalf. The user's guide's journal page covers them.

What it seesWhere the text goes once it leaves

Life area filing sends the thing being filed and your life areas with their descriptions, and gets back one area name, or the word NONE. For an entry that is its text, cut off after the first four thousand characters; for a task, its words, its kind and its note, cut at two thousand; for a note, its name, kind, description and text, cut at four thousand. Under each area it also sends up to two short snippets of things you filed there yourself and up to two of your corrections, so a line of another entry, task or note rides in the same ask as examples. Area descriptions sends your area names and the descriptions you have already written, and gets back a few sentences for the one area being drafted. Transcribing a recording sends the audio to ElevenLabs and then the transcript it returns to Anthropic. The version line sends the two versions’ text, each cut at six thousand characters, with a word saying whether the thing is a goal. The Quick Capture decision sends the words you captured. Chat sends more than any of them: your question, the recent thread, and whatever its tools then read out of your own data, which is your areas, your journal, your day, your tasks, your goals and your notes. Your photographs never leave, Chat included. Each switch in Settings states what its job sends, and Chat says what it reaches on its own guide page.

Turning it offYour switches in Settings, and the key underneath them

Your switch is under Settings → AI: a master switch called AI, and under it a switch per job. Flipping one takes effect on the next sweep, within two minutes for filing entries and tasks and for area descriptions, within five for filing notes, and within the hour for version lines, with nothing to restart. The recording and Quick Capture jobs run on the next thing you post, so a flip there is immediate.

Underneath the switches is the key everything runs on, and it is yours, in the app. Settings → Integrations holds one slot per outside service: Anthropic, ElevenLabs and Resend. A key is stored once and never shown again, and saving an empty field removes it. Every job and Chat run on the Anthropic key stored there and bill it, so removing that key stops all of them, and it is the only thing that stops Chat.

The page keeps its switches whatever the account holds, on the web and on the phone alike, so it can never lock somebody out of the settings that would fix it. With no Anthropic key stored, a line under the master switch reads No Anthropic key is stored, so nothing here runs. Add one in Settings → Integrations. Its last three words are the link to the page that fixes it. The backend's environment takes nothing away either: ANTHROPIC_API_KEY on the deployment is read at startup to fill any key slot the owner has left empty, and that is the whole of what it still does.

Turning AI back on reaches back a little for entries and all the way for the rest: the entry sweep files entries from the last two days that never got a verdict, and anything older stays as you left it, while the task and note sweeps file every open task and every note that has no area, however old.

With it offThe app keeps working, with less filled in for you

Every control still does what it says, and every screen still finishes loading.

• Entries, notes and tasks save immediately with no area filled in. You can set an area yourself, or leave it blank.
• A task under a goal still takes its goal's area. That half is arithmetic on your own data, not a model call, so it keeps working.
• A life area you have only named stays a bare name until you write what it means.
• A recording still uploads, plays and is kept. It gets no transcript, and the audio is not sent anywhere.
• A note, a goal or an event still seals as a new version. The line saying what changed is not written.
• A Quick Capture is a journal entry, every time, with the words as you typed or said them.
• Chat still answers, and still sends your question and its reading of your data to Anthropic. The master switch does not reach it. Removing the Anthropic key in Settings → Integrations is what stops it.

Everything you type is still stored, searched, and shown exactly as before.