Bucket 4: pushing to GitHub

In Lesson 1 you moved a change through three buckets on your own laptop. Now you'll send it to the fourth — the remote on GitHub — and the two words that scare beginners most, origin and push, will turn out to be boring and simple.

10-second recall (no peeking) Name the four buckets in order, and the command that moves work into each. Got it? The fourth bucket is the one we do today. (Answer: working dir → add → staging → commit → local history → push → remote.)

What a "remote" actually is

A remote is just another copy of your repository that lives somewhere else — almost always on GitHub. That's it. Your laptop has a copy (buckets 1–3). GitHub has a copy (bucket 4). They're two copies of the same project, and you keep them in sync by hand.

Why bother? Two reasons, both about fear: it's a backup (laptop dies, work survives), and it's how work is shared (other people, other machines, or just future-you).

"origin" and "main" — decoded

These aren't magic words. They're just names:

originThe nickname for your remote. "origin" is simply the default name git gives the GitHub copy. You could rename it; nobody does.

mainThe name of your branch — the line of commits you're working on. Your sandbox is on main right now.

So the command everyone copy-pastes in terror —

git push -u origin main

— reads as a plain sentence once you know the words:

Word by word git push = send my commits up to the remote · -u = and remember this pairing, so next time I can just type git push · origin = to the remote nicknamed origin · main = the main branch.

In English: "Send my main branch up to GitHub, and remember it did so."

Two directions, never confuse them again

git push ⬆️Local history → remote. Sends YOUR commits up to GitHub.

git pull ⬇️Remote → local. Brings OTHER commits down from GitHub to you.

Push = upload. Pull = download. Solo on one machine, you'll mostly push. Pull matters the moment a second machine or person touches the repo.

Try it — connect your sandbox to GitHub

Your sandbox already has a commit (bucket 3) but no remote yet. Let's give it one and push. You have the gh tool installed and logged in, so this is all from the terminal — no clicking around github.com.

# 1. go to your sandbox
cd ~/Claude-Code/learning-git/sandbox

# 2. create a PRIVATE GitHub repo and link it as "origin"
#    (--source=. means "use this folder"; no --push yet, on purpose)
gh repo create rt-git-sandbox --private --source=. --remote=origin

# 3. see the address of bucket 4 you just attached
git remote -v                 # origin  https://github.com/.../rt-git-sandbox (fetch & push)

# 4. NOW push — type it yourself, feel the words
git push -u origin main       # send main up to origin, and remember it

# 5. prove it's really on GitHub
gh repo view --web            # opens the repo in your browser

Your "add notes" commit is now in all four buckets. Now do a full lap on your own to feel the everyday loop end-to-end:

echo "learning to push" >> notes.txt   # change a file (bucket 1)
git status                             # "modified: notes.txt"
git add notes.txt                      # → staging (bucket 2)
git commit -m "Add a second line"      # → local history (bucket 3)
git push                               # → remote (bucket 4) — note: just "git push" now, thanks to -u

Refresh the GitHub page. Your second line is there. That loop — edit → add → commit → push — is the entire daily rhythm of working with git. Everything else is variations on it.

Bonus word: clone git clone <url> is the reverse of all this: it copies a remote down to a brand-new folder on your machine, buckets and all. It's how you start working on a repo that already exists on GitHub. You won't need it in the sandbox, but now the word won't surprise you.

Check yourself

From memory. Retrieval is what makes it stick.

What is origin?

origin is just the name pointing at your remote (the GitHub copy). The branch is main; the first commit has no special name.

You committed locally but never pushed. Where is your work?

A commit lives in bucket 3 (local history) until you push it to bucket 4. This is exactly why "commit" and "push" are two separate steps — and why a backup only exists after you push.

Which direction does git pull move work?

After git push -u origin main once, what can you type next time to push?

The -u ("set upstream") told git to remember the origin/main pairing for this branch. After that, a bare git push knows where to go.

I'm your teacher — ask me anything. If gh repo create or the push throws an error, paste the exact message into this Claude Code session. Errors here are usually one-line fixes, and reading them with the four-bucket picture is the fastest way to learn.

Read next (8 min): Atlassian — "Syncing: push, pull, fetch & remotes". Clear diagrams of the local↔remote sync you just did.

📄 Cheat sheet · ← Lesson 1 Next: Lesson 3 — branches, & why you're already on one