The three areas of Git

At a conceptual level, Git works with three main areas inside your project directory.​

• Working directory
  This is the normal folder where you edit, delete, and create files. It always reflects some commit plus any uncommitted changes on top.​

• Staging area (index)
  The staging area is an in-between space where you mark exactly which changes should go into the next commit. When you run git add file.js, you are telling Git: “include this version of file.js in the next snapshot.” Git stores this information in a binary file named index inside .git.​

• Repository (.git directory)
  The .git directory is a hidden folder at your project root that stores your entire history: commits, file contents, branches, tags, and configuration. If you delete this folder, Git effectively forgets that your project is a repository.​

You can think of a typical workflow like this: edit in the working directory → choose what to save using the staging area → create a permanent snapshot in the repository with git commit.​

Why the .git folder matters

The .git folder is the actual Git repository; the rest of your project is just a working copy checked out from it. This folder is created when you run git init or clone a repository.​

High-level responsibilities of .git:​

• Tracks which commit your working directory is based on.

• Stores all versions of every tracked file as compressed objects.

• Defines branches, tags, and remote-tracking branches.

• Records configuration, hooks, and reflogs for recovery.

Because everything important lives here, you can move or rename the project directory without breaking Git—as long as .git moves with it.​

Important entries inside .git

Below are the key files and subfolders you will find inside .git, along with what they do.​

Core files

• HEAD

  • A simple text file that usually contains a line like ref: refs/heads/main.

  • This means your current checkout is the branch main, and the visible code corresponds to the commit that refs/heads/main points to.​

  • In a detached HEAD state (e.g. git checkout <commit>), HEAD stores the raw commit hash instead of a ref: line.​

• config

  • Repository-local configuration, different from your global ~/.gitconfig.​

  • Typical settings include:

    • Remotes: [remote "origin"] with its fetch/push URLs.

    • Branch tracking info: which remote branch a local branch follows.

    • Repo-specific user name/email or default branch name.​

• index

  • A binary file that represents the staging area.

  • On git add:

    • Git hashes the file content, stores it as a blob in objects/, and records the path, hash, and mode in index.​

  • On git commit:

    • Git reads index to build tree objects and then a commit object, ensuring the commit exactly matches what you staged.​

• description

  • A short text description used by some Git web interfaces; it has no effect on normal CLI workflows.​

• packed-refs

  • An optimization file that stores many references (branches/tags) together when there are too many loose files under refs/.

  • Helps performance in large repositories with lots of tags or branches.​