Git Workflow

This document outlines how we manage Git at GenerativeModels.ai to ensure consistency, reliability, and speed across our development process.

---

🌳 Branching Strategy

---

🔁 Workflow Overview

1. Start a Feature

• Branch from main:

  git checkout main
  git pull
  git checkout -b feature/<short-name>
  

• Examples:

  • feature/eval-dashboard
  • feature/embed-pipeline
  • bugfix/null-user-error

---

2. Create a Pull Request (PR)

• Open a PR from feature/* to main
• Ensure:
  • ✅ Code review is done
  • ✅ Tests and linter pass via CI

Once merged, main is deployed to staging for QA and validation.

---

3. Create a Release Branch

When staging is ready for production:

git checkout main
git pull
git checkout -b release/v1.3.0

• Used to:
  • Run final production checks
  • Apply release-specific tweaks
  • Update changelogs and versioning

---

4. Production Release

When release/* is finalized:

• Merge release/* → main

• Tag the release:

  git tag v1.3.0
  git push origin v1.3.0
  

• Deploy main to production

---

5. Cherry-Pick Release Fixes to Main

If you make any commits directly on the release branch, cherry-pick them back to main to prevent divergence:

git checkout main
git cherry-pick <commit-sha>
git push origin main

---

6. Hotfix Workflow

For urgent production bugs:

git checkout main
git pull
git checkout -b hotfix/fix-payment-crash
# fix, commit, push

Then:

• Merge into main
• Tag + deploy
• Optionally cherry-pick into current release if still active

---

7. Clean Up

After merging any branch:

git branch -d feature/<n>
git push origin --delete feature/<n>

---

✅ Best Practices

• Always branch from latest main
• Keep feature branches short-lived
• Keep PRs small and focused
• Use CI, linters, and review before merging
• Tag every production release (e.g. v1.3.0)
• Don’t forget to cherry-pick changes made in release branches back to main