Notebook Cell Organizer

A VS Code extension that automatically reorganizes cells in a Jupyter Notebook — moving shell commands and imports to dedicated cells at the top.

What It Does

When working on data science notebooks, shell commands (!pip install, %pip, %conda) and import statements tend to get scattered throughout the file. This extension collects them into two clean cells at the top of the notebook:

• Cell 0 — all shell commands
• Cell 1 — all import statements

Lines are extracted from their original cells. Cells that become empty are removed. Duplicate lines are deduplicated automatically.

Before:

[cell 0]  !pip install pandas
          import numpy as np
          x = 1

[cell 1]  import pandas as pd
          df = pd.read_csv(...)

[cell 2]  import numpy as np   ← duplicate
          model.fit(X, y)

After:

[cell 0]  !pip install pandas

[cell 1]  import numpy as np
          import pandas as pd

[cell 2]  x = 1

[cell 3]  df = pd.read_csv(...)

[cell 4]  model.fit(X, y)

Usage

Command Palette

Cmd+Shift+P → Notebook: Organize Notebook Cells

Toolbar Button

Click the ⊟ button in the notebook toolbar (appears automatically when a Jupyter notebook is open).

Auto-organize on Save

Enable in settings: Cmd+, → search for organizeOnSave → toggle on.

Settings

Recognized Patterns

Shell commands (moved to cell 0):

!pip install numpy
!apt-get install build-essential
%pip install pandas
%conda install scikit-learn

Imports (moved to cell 1):

import numpy as np
from sklearn.model_selection import train_test_split

Known Limitations

• Multi-line imports with parentheses are not yet supported:

  from sklearn import (    # ← not recognized
      train_test_split,
      cross_val_score,
  )
  

• TYPE_CHECKING blocks and conditional imports are treated as regular code

Requirements

• VS Code ^1.109.0
• Jupyter extension for opening .ipynb files

License

MIT