spreadsheet

Spreadsheet Skill

When to use this skill

• Use when the task involves spreadsheets, CSV/TSV files, tabular data cleanup, or workbook analysis.
• Use when the user needs formulas, summaries, transformations, or structured exports from spreadsheet data.
• Use when the task depends on preserving headers, column types, sheet boundaries, or formulas.
• Use when bulk edits must remain auditable and reversible.

What this skill does

• Normalizes spreadsheet-like tasks into safe, repeatable tabular workflows.
• Treats headers, row counts, formulas, and delimiters as explicit assumptions rather than guessing silently.
• Prefers scripted transformations that can be reviewed and rerun.
• Helps separate data cleaning, calculation, and presentation concerns.

Implementation

• First identify the source format: CSV, TSV, XLSX, multiple sheets, exported report, or ad hoc table.
• Confirm the shape of the data before editing: headers, unique keys, empty rows, encoding, delimiters, and date formats.
• If formulas are involved, clarify whether they should be preserved, recalculated, or replaced with values.
• For bulk updates, favor scriptable transforms that produce a clear before/after story.
• Preserve source data before rewrites, especially when changing delimiters, formulas, or workbook structure.
• If workbook layout matters, call out any gap between plain-text data processing and full spreadsheet fidelity.
• Summarize row-level and column-level impacts after non-trivial edits.

Best practices

• Never assume the first row is a header without checking.
• Treat date, currency, and numeric coercion as high-risk transformations.
• Preserve formulas unless the user explicitly wants static values.
• Prefer small, auditable transformations over opaque bulk rewrites.
• State when workbook-only features such as formatting, filters, merged cells, or charts may not be preserved.