Skip to content
Get packs

Guide · Agent skill

How to use the Excel Spreadsheet Skill agent skill (xlsx)

Updated · written with AI from the project's own documentation and checked against our catalogue data

The xlsx skill is an agent skill that teaches an AI agent how to open, create, edit, fix and convert spreadsheet files such as .xlsx, .xlsm, .csv and .tsv. It is for analysts, finance staff, people cleaning messy data and developers who generate spreadsheets with Python. After this guide you will know what the skill changes about the agent's behaviour, how to install it, how to ask for spreadsheet work in plain language, and what to check before running its scripts.

Excel Spreadsheet Skilleigent-ai/eigent · Apache-2.0

Quick start

  1. 1Download and unzip. Download the xlsx ZIP from this page and unzip it so the folder contains SKILL.md at its top level.
  2. 2Place the skill. Copy the folder into ~/.claude/skills/ for all projects, or into a project's .claude/skills/ for one project. Alternatively upload it under Settings → Capabilities → Skills in the Claude apps.
  3. 3Install requirements. Make sure Python, openpyxl, pandas and LibreOffice are available. Add defusedxml if the helper scripts need it.
  4. 4Make a request. Ask for the spreadsheet task by naming the file or path, for example a column of live formulas added to sales.xlsx.
  5. 5Review the workbook. Open the result and check that calculations are formulas, the formatting is what you asked for and the inputs are clearly marked.
  6. 6Recalculate and check errors. Run python scripts/recalc.py on the file, read the JSON status, and have the agent fix any errors until the status is clean.

What the xlsx skill makes the agent do

An agent skill is a folder of instructions and optional scripts that an agent reads when a task matches the skill's description. This one applies whenever a spreadsheet file is the main input or output: editing an existing workbook, adding columns, computing formulas, formatting, charting, cleaning messy data or creating a new spreadsheet from other sources.

Once installed, the agent follows a set of rules. It writes Excel formulas instead of pasting in numbers it calculated in Python, so totals and growth rates stay live when the source data changes. For financial models it applies colour conventions: blue text for hardcoded inputs, black for formulas, green for links to other sheets and red for links to external files. It also checks the finished workbook for formula errors such as #REF!, #DIV/0! and #VALUE! by running a recalculation script.

  • Use formulas such as =SUM(B2:B9) rather than hardcoded totals
  • Keep assumptions in separate cells and reference them, for example =B5*(1+$B$6)
  • Use a consistent professional font such as Arial unless your own template says otherwise
  • Match the existing style of a template you ask it to update

Install the xlsx skill from the ZIP or the Claude apps

Download the ZIP from this page and unzip it. To make the skill available to every project on your machine, place the unzipped xlsx folder inside ~/.claude/skills/. To limit it to one project, place it inside that project's .claude/skills/ folder instead.

If you use the Claude apps, you can upload the skill instead. Open Settings, go to Capabilities, then Skills, and upload the skill there. Check that the folder contains SKILL.md at its top level, since that file is what the agent reads to decide when to use the skill.

  • Personal install: ~/.claude/skills/
  • Project install: .claude/skills/ inside the project folder
  • Claude apps: Settings → Capabilities → Skills

Install the requirements the scripts need

The skill expects Python, openpyxl and pandas for reading and writing workbooks. Formula recalculation uses scripts/recalc.py, which needs LibreOffice installed. The bundled Office helper scripts use defusedxml. The source says the skill can assume LibreOffice is present and that recalc.py configures it on first run, including in sandboxed environments where Unix sockets are restricted.

If a script fails with a missing module or program, install the missing item and try again. The source does not give specific install commands for these packages, so use the official instructions for your operating system.

  • Python
  • openpyxl
  • pandas
  • LibreOffice (required for recalc.py)
  • defusedxml (used by the bundled Office helper scripts)

Trigger the skill with natural requests

You do not need to name the skill. Refer to the spreadsheet by file name or path, even casually, and say what you want done to it or produced from it. The skill is also meant to trigger for cleaning messy tabular files with malformed rows, misplaced headers or junk data.

The skill is designed not to trigger when the main deliverable is a Word document, HTML report, standalone Python script, database pipeline or Google Sheets API integration. If you want one of those, ask for it directly so the agent does not pick the spreadsheet workflow.

  • Add a Margin % column to the workbook in my downloads called sales.xlsx, using formulas
  • Clean up this messy CSV: the header is on row 4 and there are blank rows to remove, then save as a proper .xlsx
  • Convert budget.tsv to an Excel file with the totals as live formulas
  • Fix the #DIV/0! errors in forecast.xlsx and recalculate

Check the output and run the recalculation script safely

After the agent saves a workbook that contains formulas, the skill's workflow is to run recalculation with scripts/recalc.py and read the JSON it returns. If status is errors_found, the error_summary lists the error types and their locations, and the agent should fix them and recalculate again. Ask the agent to confirm the final status before you rely on the file.

Recalc.py runs code and LibreOffice on your machine. The source warns against recalculating sensitive workbooks on a shared machine without reviewing the scripts first. The skill does not use the network, according to this listing, but read the scripts in the folder before running them on anything confidential.

  • Review scripts/recalc.py and scripts/office/ before first use on sensitive data
  • Keep a copy of the original workbook before the agent edits it
  • Look for status errors_found and the error_summary in the JSON output
python scripts/recalc.py output.xlsx

What people use it for

  • Adding formula-driven columns such as margins, growth rates and totals to an existing workbook
  • Building a financial model with blue inputs, black formulas and assumption cells
  • Cleaning a messy CSV or TSV with misplaced headers and blank rows into a formatted .xlsx
  • Converting tabular data between CSV, TSV and Excel formats
  • Generating spreadsheets from Python data with pandas and openpyxl

Frequently asked questions

Is the xlsx agent skill free?

The listing on this page says the repository is Apache-2.0, but the SKILL.md front matter says Proprietary and LICENSE.txt restricts copying, redistribution and derivative works except under Anthropic's service agreements. Read LICENSE.txt yourself before reuse, and do not assume the Apache-2.0 label covers this skill's files.

What do I need installed to use the xlsx skill?

The source lists Python, openpyxl and pandas for reading and writing workbooks, and LibreOffice for formula recalculation with scripts/recalc.py. defusedxml is used by the bundled Office helper scripts. The skill runs scripts, so you should be comfortable reviewing and running them.

Where do I put the xlsx skill folder?

Put it in ~/.claude/skills/ to make it available across projects, or in a project's .claude/skills/ folder to limit it to that project. In the Claude apps you can instead upload it under Settings → Capabilities → Skills.

Why does the agent recalculate my file and report errors?

The skill requires zero formula errors and asks the agent to run scripts/recalc.py after saving formulas. The script returns JSON, and if status is errors_found the error_summary shows where #REF!, #DIV/0!, #VALUE! or #NAME? appear so they can be fixed.

Can the xlsx skill convert CSV or TSV files to Excel?

Yes. The description covers converting between tabular formats and opening, editing and fixing .xlsx, .xlsm, .csv and .tsv files. Pandas can read and write Excel files, and the skill's examples use df.to_excel with index=False.

How is this different from the Word or PowerPoint helpers?

The bundled scripts/office helpers cover Word and PowerPoint files, which this spreadsheet skill mostly does not use. For spreadsheet work the important parts are the formula rules, formatting conventions and recalc.py.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

Choose the xlsx skill if you want an agent to build or repair workbooks with live formulas, consistent formatting and a recalculation check, or to clean messy CSV and Excel data. Pick something else if your main output is a Word document, HTML report, database pipeline or Google Sheets integration, which the skill itself excludes, or if you cannot review scripts that run code on sensitive files.

Download the skill