First Run and Onboarding

The first time you open Wayland, it walks you from a blank state to a working assistant. The whole flow takes about a minute.

1. Sign in

The app opens on a login screen. Sign in to create your local profile. Your data stays on your machine; the account ties your settings and memory together so they follow you across devices if you later turn on sync.

2. Connect a provider

Onboarding asks you to connect at least one model provider so the agent has something to think with. You have two ways to authenticate:

• Anthropic OAuth: sign in with your Claude account and use your existing subscription. No API key to paste.
• API key: paste a key for any supported provider (Anthropic, OpenAI, Google, and many more).

You can add more providers later in Settings -> Models. One is enough to begin. For a focused walkthrough, see Connect Your First Model.

3. Memory setup

Wayland offers to set up its memory backbone (IJFW Memory) so assistants can remember decisions and context across sessions. You can accept it now or skip it and enable it later from Settings -> IJFW Memory. Skipping does not block anything else.

4. Land on the dashboard

When onboarding finishes you arrive at the dashboard: a launch bar with your assistants, a set of intent pills (Sell, Write, Research, Build, Run, Plan, Learn), and starter cards. This is home base for everything that follows.

Desktop vs Core: which do I use

Now that the app is running, it helps to know when to drop to the engine:

• Use the desktop app for channels, teams, scheduling, voice, image generation, and remote access. It is the full surface.
• Use Wayland Core for headless work: a one-shot prompt in a script, an agent in CI, or driving the agent from your own program. See Wayland Core Overview.

They share config and run the same engine, so moving between them is seamless.

Next

• Connect Your First Model
• Your First Conversation