notebook.split_header_cells

Goal

• Split each markdown cell in a Jupyter notebook so it has at most one header

Workflow

Split Markdown Cells

• Make sure each markdown cell has at most one header and some text
  • Bad (there are 3 headers in the same cell: one H1, one H2, one H3)

    # %% [markdown]
    # Part 3: Composition Examples
    
    ## Example 1: Minimal End-to-End Workflow
    
    Rain -> Sprinkler -> Grass Wet
    
    ### Mental Model
    

  • Good (each header is in a different cell)

    # %% [markdown]
    # Part 3: Composition Examples
    
    # %% [markdown]
    ## Example 1: Minimal End-to-End Workflow
    
    Rain -> Sprinkler -> Grass Wet
    
    # %% [markdown]
    ### Mental Model
    

Remove Empty Lines

• Remove empty lines at the start or end of a markdown cell
  • Bad (blank lines around the header)

    
    # %% [markdown]
    # Part 3: Composition Examples
    
    
    

  • Good (no blank lines around the header)

    # %% [markdown]
    # Part 3: Composition Examples
    

Sync with Jupytext

• At the end, run jupytext --sync to update the paired Python notebook, per # Code Architecture and Responsibility -> ## Utilities vs. Notebook Responsibilities in .claude/skills/notebook.rules.md

Constraints

• Do not change or remove any Python code cell
• Do not change the content of the markdown text besides splitting cells

Verification

• Confirm no markdown cell contains more than one header
• Confirm every Python code cell is unchanged
• Confirm the .ipynb and paired .py file are in sync via Jupytext