Two GitHub accounts, one computer · step 7 of 7

A new repository and a new device

How to start a repo from scratch, connect to an existing one on a new machine, and pull the latest data.

This step assumes you already have an account configured (SSH key + .gitconfig-userX) following the earlier steps in this tutorial. Below are three separate scenarios — pick the one you need.

A. A brand-new repository from scratch

1. Create an empty repository on GitHub, on the user1 account (the New repository button — don’t check “Initialize with README”, to avoid a conflict on the first push). Copy the SSH address GitHub shows on the next screen — it looks like git@github.com:user1/repo-name.git.

2. Create the project folder in the right place (i.e. wherever you have includeIf configured for user1):

cd D:/Projects/user1
mkdir repo-name
cd repo-name

3. Initialize the repository locally:

git init

4. Check that Git actually sees the user1 identity (thanks to includeIf from the earlier steps):

git config --show-origin --get user.email

It should show user1@example.com from .gitconfig-user1. If it shows something else, go back to the “Configuring .gitconfig” step before continuing.

5. Wire up the remote (the address you copied in step 1) and make your first commit:

git remote add origin git@github.com:user1/repo-name.git
echo "# Project name" > README.md
git add README.md
git commit -m "Initial commit"

6. Push to GitHub:

git branch -M main
git push -u origin main

The -u flag remembers the link between your local main branch and the remote one — from now on, plain git push is enough.

B. Connecting to an existing repo on a new device

We’re assuming that on the new machine you go through steps 1–2 of this tutorial from scratch (a new SSH key pair, a new .gitconfig, includeIf, adding the public key to GitHub) — SSH keys don’t “carry over” by themselves; you either copy them securely from the old machine, or (simpler and safer) generate a fresh pair and add it as another key on the same GitHub account.

1. Confirm the setup works before attempting to clone:

ssh -T -i path/to/key -o IdentitiesOnly=yes git@github.com

It should show Hi user1!.

2. Clone the repository into the right folder (the one that has includeIf configured):

cd D:/Projects/user1
git clone git@github.com:user1/repo-name.git

Git will use SSH right away (since the URL starts with git@github.com:), so it’ll land on the correct account automatically — as long as the key for that folder is already wired up in core.sshCommand.

3. Check the identity in the freshly cloned repo:

cd repo-name
git config --show-origin --get user.email

C. Pulling the latest changes (an existing, already cloned folder)

If you already have the repo locally and want to fetch changes someone (or you, from another device) pushed in the meantime:

cd D:/Projects/user1/repo-name
git pull

git pull is really git fetch (download changes from GitHub) plus git merge (blend them into your local branch) in one command. If you’d rather see what changed before merging it into your work:

git fetch
git log HEAD..origin/main --oneline

This second command lists the commits that exist on GitHub but not yet locally, without pulling them into your current branch.

If git pull reports a conflict with your uncommitted changes, run git status first to see what’s unsaved, then either git stash (set your changes aside) or git commit (save them) before trying again.