xlsx

---
name: xlsx
description: Use when a spreadsheet is the primary input or output — reading, editing, creating, or converting .xlsx, .xlsm, .csv, or .tsv files, including cleaning messy tabular data.
source: https://github.com/anthropics/skills/tree/main/skills/xlsx
generated: 2026-07-02T18:43:53.484Z
category: tool
audience: general
---

## When to use

- Opening, reading, editing, or fixing an existing .xlsx, .xlsm, .csv, or .tsv file
- Creating a new spreadsheet from scratch or from other data sources
- Cleaning or restructuring messy tabular data — malformed rows, misplaced headers, junk — into a proper spreadsheet
- Building a financial model where the deliverable is a spreadsheet with live formulas and correct formatting

## Key concepts

### pandas vs openpyxl

pandas is best for data analysis, bulk operations, and simple export; openpyxl is best for formulas, formatting, and Excel-specific features. The skill picks the tool per task.

### Formulas, not hardcoded values

Always write Excel formulas (=SUM(B2:B9)) instead of computing in Python and hardcoding the result, so the spreadsheet stays dynamic and recalculates when source data changes.

### scripts/recalc.py recalculation

openpyxl writes formulas as strings without values; recalc.py drives LibreOffice to recalculate all sheets and returns JSON reporting error types and locations.

### Zero formula errors requirement

Every model must ship with no #REF!, #DIV/0!, #VALUE!, #N/A, or #NAME? errors — verified by scanning recalc.py output and fixing until clean.

### Financial model formatting standards

Industry color coding (blue inputs, black formulas, green intra-workbook links, red external links, yellow key assumptions) plus number-format rules for currency, percentages, multiples, and negatives.

### Preserve existing templates

When modifying an existing file, exactly match its established format and conventions; the template's patterns always override the skill's default guidelines.

## API reference

```
npx skills add anthropics/skills --skill xlsx
```

Install the spreadsheet creation, editing, and analysis skill.

```
npx skills add anthropics/skills --skill xlsx
```

```
python scripts/recalc.py <excel_file> [timeout_seconds]
```

Recalculate all formulas via LibreOffice and report any Excel errors as JSON.

```
python scripts/recalc.py output.xlsx 30
```

## Gotchas

- recalc.py needs LibreOffice installed; formulas written by openpyxl have no values until it runs
- Opening with data_only=True and saving permanently replaces formulas with their cached values — data loss
- openpyxl cell indices are 1-based (row=1, column=1 is A1); mixing with 0-based pandas rows causes off-by-one bugs
- Never impose standardized formatting on a file with an established template; its conventions override the defaults
- Don't hardcode calculated results in Python — use Excel formulas so the sheet recalculates on data changes
- Don't trigger this skill when the real deliverable is a Word doc, HTML report, standalone script, or Google Sheets integration

---
Generated by SkillMake from https://github.com/anthropics/skills/tree/main/skills/xlsx on 2026-07-02T18:43:53.484Z.
Verify against source before relying on details.

File: ~/.claude/skills/xlsx/SKILL.md