// Docs

Getting started.

The complete walkthrough, every click and every command, lives on the get-started page. It was written for people who have never opened a terminal. If you are installing right now, go there.

This page is the map. Setup is three acts.

Act one: install

One script does the heavy lifting. Your invitation email carries it, along with your personal key. Save the attachments into Documents, run the script, and wait for it to say it is done. It installs Claude Code and the Klyr plugin. If it gets interrupted, run it again: it picks up where it left off.

Act two: /klyr:init

Open Claude Code and type /klyr:init. It is a guided setup that asks one simple question at a time:

• Where your clone's memory lives. Hosted (we run it, works on any machine) or local (your machine, free, needs a strong computer). The one real choice of the setup.
• Its name. Pick one or let it name itself.
• KYH, the interview. Optional now, available anytime later with /klyr:kyh.
• Your workspace. A plain folder where your clone files what you share.

Nothing is final and nothing breaks. /klyr:init is safe to re-run: it remembers where you were and continues.

Act three: first sessions

Just talk to it. Tell it what you are working on, what you decided, what is bugging you. It remembers, and every session it starts a little less generic. Ask it "what do you know about me?" after a week. That is usually the moment it clicks.

Next: Your clone's memory.