XLSX Skill

Overview

XLSX Skill is one of Anthropic's official Claude Skills, designed for comprehensive spreadsheet creation, editing, and analysis with full support for formulas, formatting, data analysis, and visualization. This skill enables Claude to work professionally with Excel files (.xlsx, .xlsm) and related formats (.csv, .tsv), handling everything from simple data operations to complex financial models.

The skill emphasizes using Excel formulas instead of hardcoded values, ensuring spreadsheets remain dynamic and updateable. It includes extensive quality standards for financial models, automatic formula recalculation, and comprehensive error detection.

Core Features

1. Spreadsheet Creation

Create professional Excel files with:

• Formulas: Dynamic calculations that update automatically
• Formatting: Fonts, colors, alignment, borders
• Structure: Multiple sheets, tables, charts
• Styles: Cell styles and conditional formatting
• Rich Content: Images, hyperlinks, comments

2. Data Analysis

Analyze spreadsheet data using pandas:

• Read and process Excel files
• Statistical analysis and aggregation
• Data transformation and cleaning
• Multi-sheet operations
• Export to various formats

3. Formula Recalculation

Automatic formula calculation with LibreOffice:

• Recalculate all formulas in all sheets
• Comprehensive error detection (#REF!, #DIV/0!, #VALUE!, etc.)
• Detailed error reporting with locations
• JSON output for automated processing

4. Financial Model Standards

Professional financial modeling conventions:

• Color Coding: Blue for inputs, black for formulas, green for links
• Number Formatting: Currency with units, zeros as dashes, parentheses for negatives
• Formula Construction: Assumptions in separate cells, documented hardcodes
• Error Prevention: Zero formula errors required

5. Quality Assurance

Comprehensive verification and validation:

• Formula error detection
• Cell reference verification
• Division by zero checks
• Cross-sheet reference validation
• Edge case testing

Use Cases

• Financial Models: Build dynamic financial projections and analyses
• Data Analysis: Analyze and transform large datasets
• Reports: Generate formatted business reports
• Budgets: Create and track budgets with formulas
• Dashboards: Build data visualization dashboards
• Data Import/Export: Convert between Excel and other formats

Technical Implementation

Creating Spreadsheets

Uses openpyxl for formulas and formatting:

from openpyxl import Workbook
from openpyxl.styles import Font, PatternFill

wb = Workbook()
sheet = wb.active
sheet['A1'] = 'Revenue'
sheet['B1'] = '=SUM(B2:B10)'  # Formula, not hardcoded value
sheet['A1'].font = Font(bold=True, color='0000FF')
wb.save('output.xlsx')

Analyzing Data

Uses pandas for data operations:

import pandas as pd

df = pd.read_excel('file.xlsx')
df.describe()  # Statistical analysis
df.to_excel('output.xlsx', index=False)

Recalculating Formulas

Uses LibreOffice via recalc.py script:

python recalc.py output.xlsx

Returns JSON with error details and locations.

Critical Requirements

Use Formulas, Not Hardcoded Values

Always use Excel formulas instead of calculating in Python:

❌ WRONG - Hardcoding:

total = df['Sales'].sum()
sheet['B10'] = total  # Hardcodes 5000

✅ CORRECT - Using formula:

sheet['B10'] = '=SUM(B2:B9)'  # Dynamic formula

Zero Formula Errors

Every Excel file must be delivered with ZERO formula errors:

• #REF! (invalid references)
• #DIV/0! (division by zero)
• #VALUE! (wrong data type)
• #N/A (not available)
• #NAME? (unrecognized name)

Use the recalc.py script to verify and fix all errors.

Financial Model Standards

Color Coding

Industry-standard conventions:

• Blue text: Hardcoded inputs users will change
• Black text: All formulas and calculations
• Green text: Links from other worksheets
• Red text: External file links
• Yellow background: Key assumptions needing attention

Number Formatting

Professional formatting rules:

• Years: Text strings ("2024" not "2,024")
• Currency: $#,##0 with units in headers ("Revenue ($mm)")
• Zeros: Display as "-" including percentages
• Percentages: 0.0% format (one decimal)
• Negatives: Parentheses (123) not minus -123

Formula Construction

Best practices:

• Place assumptions in separate cells
• Use cell references instead of hardcoded values
• Document all hardcoded data sources
• Verify all cell references
• Test with edge cases

Workflow

Standard Process

1. Choose tool: pandas for data, openpyxl for formulas/formatting
2. Create/Load: Create new or load existing file
3. Modify: Add data, formulas, formatting
4. Save: Write to file
5. Recalculate (MANDATORY): python recalc.py output.xlsx
6. Verify: Fix any errors and recalculate again

Formula Verification Checklist

• Test 2-3 sample references first
• Verify column mapping (column 64 = BL)
• Check row offset (DataFrame row 5 = Excel row 6)
• Handle NaN values with pd.notna()
• Check denominators for division by zero
• Verify cross-sheet references
• Test edge cases

Error Detection

The recalc.py script returns comprehensive error information:

{
  "status": "errors_found",
  "total_errors": 2,
  "total_formulas": 42,
  "error_summary": {
    "#REF!": {
      "count": 2,
      "locations": ["Sheet1!B5", "Sheet1!C10"]
    }
  }
}

Fix identified errors and recalculate until status is "success".

Best Practices

Library Selection

• pandas: Data analysis, bulk operations, simple export
• openpyxl: Formulas, formatting, Excel-specific features

Working with openpyxl

• Cell indices are 1-based (row=1, column=1 = A1)
• Use data_only=True to read values (but don't save - loses formulas!)
• Formulas preserved but not evaluated until recalc.py runs
• For large files: read_only=True or write_only=True

Working with pandas

• Specify data types to avoid inference issues
• Read specific columns for large files
• Handle dates properly with parse_dates

Dependencies

• openpyxl: Excel file manipulation
• pandas: Data analysis
• LibreOffice: Formula recalculation
• defusedxml: Secure XML parsing

Summary

XLSX Skill enables Claude to create professional, dynamic spreadsheets with formulas, formatting, and comprehensive quality assurance. Through industry-standard conventions, automatic formula recalculation, and zero-error requirements, this skill ensures spreadsheets are reliable, maintainable, and professionally constructed for business, finance, and data analysis use cases.