Data Manuscript

Drag in a spreadsheet: the workbook splits into tabs, column types are inferred, charts appear immediately, and the data never leaves the page.

Loading preview...
Start with this prompt

Drag in a spreadsheet: the workbook splits into tabs, column types are inferred, charts appear immediately, and the data never leaves the page.

Try Websites Agent
Build a local spreadsheet analysis page called Data Manuscript, one that turns a raw file into a readable manuscript of charts. When the visitor drags in a CSV or XLSX file, the file is parsed entirely inside the browser, and the page states clearly that the data never leaves this page. A workbook with several sheets is automatically split into tabs, one tab per sheet, and the field type of every column is inferred as number, date or category. Numeric columns are charted directly, with distribution, trend and proportion views, while category columns become filters that link across all the charts on the page. An empty or unreadable file produces a clear error message instead of a broken view, and a built-in sample dataset can be loaded with one click so the whole flow can be tried immediately. Keep the whole page in a clean, quiet document style. Single-file static page.
Add a pivot table

Adds a pivot table: pick any dimension and measure to generate a cross summary table that works alongside the automatic charts and the linked filters, with the data still parsed only in the browser.

Try Websites Agent
Build a local spreadsheet analysis page called Data Manuscript, one that turns a raw file into a readable manuscript of charts. When the visitor drags in a CSV or XLSX file, the file is parsed entirely inside the browser, and the page states clearly that the data never leaves this page. A workbook with several sheets is automatically split into tabs, one tab per sheet, and the field type of every column is inferred as number, date or category. Numeric columns are charted directly, with distribution, trend and proportion views, while category columns become filters that link across all the charts on the page. An empty or unreadable file produces a clear error message instead of a broken view, and a built-in sample dataset can be loaded with one click so the whole flow can be tried immediately. Keep the whole page in a clean, quiet document style. Single-file static page.
Flag outlier values

Numeric charts now flag the data points that deviate sharply from the rest, so suspicious values stand out on the distribution, trend and proportion views at a glance.

Try Websites Agent
Build a local spreadsheet analysis page called Data Manuscript, one that turns a raw file into a readable manuscript of charts. When the visitor drags in a CSV or XLSX file, the file is parsed entirely inside the browser, and the page states clearly that the data never leaves this page. A workbook with several sheets is automatically split into tabs, one tab per sheet, and the field type of every column is inferred as number, date or category. Numeric columns are charted directly, with distribution, trend and proportion views, while category columns become filters that link across all the charts on the page. An empty or unreadable file produces a clear error message instead of a broken view, and a built-in sample dataset can be loaded with one click so the whole flow can be tried immediately. Keep the whole page in a clean, quiet document style. Single-file static page.
Export charts as images

Every chart generated from a numeric column can be exported as a PNG or SVG image for saving and reuse; the automatic charting and linked filtering flow is unchanged.

Try Websites Agent
Build a local spreadsheet analysis page called Data Manuscript, one that turns a raw file into a readable manuscript of charts. When the visitor drags in a CSV or XLSX file, the file is parsed entirely inside the browser, and the page states clearly that the data never leaves this page. A workbook with several sheets is automatically split into tabs, one tab per sheet, and the field type of every column is inferred as number, date or category. Numeric columns are charted directly, with distribution, trend and proportion views, while category columns become filters that link across all the charts on the page. An empty or unreadable file produces a clear error message instead of a broken view, and a built-in sample dataset can be loaded with one click so the whole flow can be tried immediately. Keep the whole page in a clean, quiet document style. Single-file static page.