This is an intimate and kind of weird relationship with an LLM. Don't overanthropomorphize it, and don't think about it too hard.
This is a loose kit to get you started, not a straitjacket.
My operating assumption in writing it is that you have ADHD, you're kind of a mess, you're aware you need better structure and better organization, but you're not really sure what that looks like for you. So xo is going to make some assertions to fill that presumed gap.
When you go, “No, I want to do it my way instead,” do it with my explicit blessing.
Ultimately, the only way this works is if you build it bespoke for yourself, around yourself.
xo starts with a message you paste into Claude Code or Codex, not ordinary chat. Those are the parts of the apps that can work with files on your computer. You don't need to be a developer or have a coding task.
Start with something useful: draft an appointment request, turn a scattered to-do list into one next action, or pick up an unfinished task without rebuilding it from memory. Stopping after a useful result is fine.
Some brief terminology: You're Hero. XO is your agent. xo is this larger package.
From there you can:
Saving your place doesn't commit you to the full setup. Connecting email or a calendar is optional; this does not start monitoring them automatically.
After useful work, XO can offer to keep going and make your XO work better for you. You decide whether you have the time and energy. It can also notice something that keeps getting in the way and suggest one useful improvement, explaining the benefit, effort, cost and limits before you choose. Saying no is fine; it won't keep asking or change its instructions without showing you the change and waiting for your yes.
xo is for people who have used chatbots but are new to assistants working with files on their computer.
Already have a setup? Browse and borrow. Read for ideas, keep what helps, and adapt it to what you already use. You don't need to install xo or replace your setup.
Optional: Kettle: You know a guy? A separate address book for finding someone you already know. Start with one contact export you choose, in a private store outside xo. Research and a setup interview are optional; you don't need Kettle to use xo. Before connecting, know that its reader can search the whole index, not just the person you ask about. Reads through a cloud model expose the approved content to its provider; local files alone don't establish verified local model processing.
This is not a code framework.
This is not an app.
This is not a larger holistic system for running your life.
It will not itself run errands for you or apply to jobs for you or email your mother.
It will, inshallah, provide a logistical framework that helps you do all of those things.
With a cloud app, everything you say or show XO goes to the company running its model. For a local app, check cloud routing, fallback and telemetry before assuming your conversations stay on your computer.
XO will help you find the provider's training setting when you get started. Turning training off does not erase what you've already sent or necessarily shorten how long the provider keeps it.
XO's memory uses plain and portable files on your computer by default. New memory can start with one small file containing something you choose to keep. As finding the right context gets harder, XO can propose separate files and indexes for more precise recall. You approve what gets saved and each change; an existing organized setup stays intact. You can choose the fuller structure from the start, or save nothing. An optional private GitHub backup puts a copy on GitHub's servers; you choose whether to use it after reviewing what would leave your computer.
You'll need access to Claude Code or Codex in the desktop app; check your existing plan before spending anything to test this.
Start with a task here. Saving your place takes some setup; connecting other apps is optional and can wait. A note saves the result. A separately chosen return setup aims to let you pick the familiar project folder and recover your chosen work in a new conversation, without keeping a prompt. XO checks that it works in your app before calling it ready, and never resumes paused installation without your choice. XO should keep your chosen task in view and return to it after a detour; that does not depend on you saying “back to my task.” You can say it whenever it helps. If you'd like more help stopping, ask for a cue after one task or before another setup change. You can choose to keep some friction; XO is not a timer or an activity monitor.
The apps have usage limits and we might max out a five-hour window while setting up. If we do, that'll be annoying and you'll have to pick up later.
AI makes mistakes and you'll have to correct it sometimes; I feel that these are valuable opportunities to demonstrate continued human relevance but they are, also, very annoying.
This is brand new and not everything is tested yet. Anyone who wants to try it is welcome; tell me where you get stuck.
Pick Claude or ChatGPT, open Code or Codex in its desktop app, and paste the starting message.
You can switch models, apps, or both later and keep your saved instructions and memory files. A new app still needs its own setup and connections; don't assume it carries over your current conversation.