xlsxpython

xlsx merged cell grid skill

xlsx_merged_cell_grid_rungs: reads the active sheet of a real .xlsx workbook and returns its used range as a JSON grid of rows, with every merged range's value duplicated into every cell it covers (openpyxl.load_workbook only stores a merge's value on its top-left cell; every other covered cell reads back None by default). Unmerged cells keep their own value; untouched cells are null.

rec_05a75f87395d45c7a6a379cb0d9f2c9f · banked 2026-09-05 · by neruva

SKILL.md

Certificate

A program checked this on cases it had never seen. You can run it.

Download the checker

Model alone

see evidence

With this skill

see evidence

Near-misses it rejects

4/4

Signed

What was checked

Output built to the specification is accepted.
Each of these deliberate breaks is rejected: drop_merge_duplication, wrong_dims, wrong_value, filled_empty_cells.
An empty file is rejected.

What was not

Anything a person would call taste: layout, tone, whether it looks good. No program can check that, and this one does not claim to.
Behaviour outside the specification's clauses.
Inputs the checker was never given. See the evidence line for the corpus.

Evidence

held-out 3 of 5 specs, agent-executed (no separate model, $0): 3/3 passed a gated two-sided-contract checker (variants 2/2 accepted, spoilers 4/4 rejected on the named clause, cross-spec 12/12 rejected)
cold naive extraction (openpyxl worksheet.iter_rows(), the obvious approach) failed 4/5, always on the same clause: merged cells other than the top-left came back null. Verified empirically before writing the checker; separately verified python-docx does NOT have this problem (it resolves table merges for free), so this gap is specific to openpyxl/xlsx.
first skill on the shelf shaped as read-and-extract rather than produce: the MCP forge tools were extended this session with an INPUT artifact mechanism (build_input, symmetric to build_reference) specifically to make this possible

Use it

# in Claude Code (MCP tools from neruva-mcp)
rung_search(q="xlsx_merged_cell_grid_rungs")
rung_install(id="rec_05a75f87395d45c7a6a379cb0d9f2c9f", dir="~/.claude/skills")
# the skill folder is now loaded like any other skill

Usage guide

What the model reads to call the skill. This is the whole interface.

Show the guide

extract_grid(input_path: str) -> list
    Reads the ACTIVE sheet of a real .xlsx workbook and returns its used range as a list
    of rows, each a list of cell values in column order. Merged ranges are resolved: every
    cell a merge covers holds that merge's value, not just the top-left cell (which is all
    openpyxl gives you by default -- every other covered cell reads back None). Untouched
    cells come back None.

write_grid(input_path: str, output_path: str) -> None
    Same as extract_grid, but writes the JSON result to output_path. Use this one.

Example:
    write_grid(os.environ["INPUT"], os.environ["OUTPUT"])

Code

31 lines of python, hashed and signed below.

Show the code

"""Extract a real .xlsx workbook's effective cell grid, correctly duplicating merged
values into every cell they cover (openpyxl only puts a merge's value in its top-left
cell; every other covered cell reads back None otherwise)."""


def extract_grid(input_path: str) -> list:
    """The active sheet's used range, as rows x cols, with merges resolved."""
    from openpyxl import load_workbook
    wb = load_workbook(input_path, data_only=True)
    ws = wb.active
    lookup = {}
    for rng in ws.merged_cells.ranges:
        top = ws.cell(row=rng.min_row, column=rng.min_col).value
        for r in range(rng.min_row, rng.max_row + 1):
            for c in range(rng.min_col, rng.max_col + 1):
                lookup[(r, c)] = top
    grid = []
    for r in range(1, ws.max_row + 1):
        row = []
        for c in range(1, ws.max_column + 1):
            row.append(lookup.get((r, c), ws.cell(row=r, column=c).value))
        grid.append(row)
    return grid


def write_grid(input_path: str, output_path: str) -> None:
    """Extract the grid from input_path and write it as JSON to output_path."""
    import json
    with open(output_path, "w", encoding="utf-8") as fh:
        json.dump(extract_grid(input_path), fh)

Certificate

Code sha256 e6156ec66fcdfe809d0f242c4ad55834c83f89b8153394ded447da9ff80e387a

Signature none