Git cheat sheet for data people

git config --global user.name "Your Name"
git config --global user.email "you@example.com"
git config --global init.defaultBranch main

git init                         # new repository
git clone https://github.com/org/repo.git

git status                       # what changed?
git add models/fct_orders.sql    # stage one file
git add -p                       # stage piece by piece
git commit -m "Add fct_orders model"
git push
git pull --rebase                # get teammates' work

git switch -c feature/orders-model   # create and move
git switch main
git branch -d feature/orders-model   # delete merged branch
git merge feature/orders-model

git log --oneline --graph --all
git diff                     # unstaged changes
git diff --staged            # staged changes
git blame models/dim_customers.sql
git show <commit>

git restore file.sql             # discard local edits
git restore --staged file.sql    # unstage
git commit --amend               # fix last commit message
git revert <commit>              # undo a pushed commit
git reset --soft HEAD~1          # undo last local commit

git stash push -m "wip"
git stash list
git stash pop

.venv/
__pycache__/
.env
target/
dbt_packages/
logs/
*.xlsx

git switch main && git pull
git switch -c fix/revenue-definition
# ...edit, then:
git add -p && git commit -m "Exclude refunds from revenue"
git push -u origin fix/revenue-definition
# open a pull request, let CI run dbt build, get a review, merge