Onboarding

my has two setup-shaped commands with different jobs:

• my onboarding is the guided first-run experience. In an interactive terminal it launches a harness and starts a learn-by-example walkthrough. The operator opens a second terminal or split pane, runs small sets of validated my commands, and confirms each set before the model continues.
• my onboarding --no-agent is the deterministic walkthrough. It explains the model and points at my setup --interactive without launching a harness.
• my setup is the machine configurator. It remains deterministic and safe for scripts; add --interactive only when you want prompts.

my onboard remains available as a compatibility alias.

First Run

my onboarding

When a single logged-in harness is detected, my launches it. If none is logged in but exactly one supported harness is installed, my launches that harness. If the choice is ambiguous, my asks which harness to use. Pass --harness codex, --harness claude-code, --harness opencode, or --harness antigravity to skip detection.

The launched model starts by greeting the operator and immediately sets up the split-pane workflow. It gives a small first command set, waits for the operator to confirm whether it worked, and then continues one set at a time. The normal path is conversational: ask whether the command worked, answer questions, and move on. If something fails or the operator is unsure, the model can offer read-only checks such as my doctor, but verification is support, not a hard gate.

The walkthrough detects whether this machine is authoring a new organization or joining an existing one:

• With no registered manifest, the harness starts from the current directory and takes the AUTHOR branch. The model interviews the operator, asks for approval before presenting my init, then walks through the smallest useful local setup and verification path: my setup, my doctor, harness launch, work sessions, and sync.
• With a registered manifest, the launcher reuses the normal my ai --setup --no-session path and takes the JOIN branch. The model helps pick a role when needed, has the operator run setup, pulls workspace content, and teaches the basic daily loop: launch a harness, start/resume/finish a work session, run pull-only my sync, preview publish with my sync --push --print, publish with my sync --push, and run my doctor.

Onboarding deliberately avoids teaching the full CLI. For meeting transcripts, fleet/support context, notes, screenshots, or issue details, the human should paste the raw context into the harness chat. Agents are the primary operators and can choose the right deeper my commands when records need to be created.

If a selected launch skill already exists from a manual install or old workspace rename, interactive onboarding asks whether to replace or skip that entry and then continues launching the harness.

The launcher itself does not publish anything. The onboarding guidance keeps publish at the end: the model must have the operator run my publish --print, review the planned remotes and pushes, and get explicit human approval before the real my publish.

Deterministic Walkthrough

Use the non-agent path in scripts, CI, or terminals where you only want the fixed setup review:

my onboarding --no-agent

If no manifest is registered, the walkthrough prints the registration command to run once you have the manifest URL:

my manifests add <name> <git-url>

No umbrella-local tour state is written until setup has created or loaded the umbrella. Once a manifest is available, the walkthrough explains the control plane and data plane, shows what setup will change, and offers to run:

my setup --interactive

Reconfigure Later

Use setup directly:

my setup --interactive

The interactive path can choose among registered manifests and select a role. Enter none at the role prompt to clear the selected role and return to unscoped guidance/services. Plain my setup never prompts, including on a TTY.

Repeat Onboarding

Run my onboarding again when you want the model to re-introduce and review the workspace with the operator. Use my onboarding --no-agent for a read-only status review that reports the umbrella, selected role, and next commands without silently re-running setup.