Overview

excel and googleSheets are the main structured-data entry points in workflows. They typically output arrays of row objects, which makes them the best source nodes for field mapping and variable expressions.

Configuration / Settings

Node	Setting	Type	Required	Details
`excel`	`operation`	`read`, `write`, `transform`, `merge`, `split`	No	The sidebar currently exposes `read` and `write`; the executor supports more operations.
`excel`	`sheetName`	string	No	Optional worksheet selector.
`googleSheets`	`connectionId`	string	Yes	Required for API access.
`googleSheets`	`operation`	`read`, `write`, `append`, `clear`	No	The panel exposes `read`, `write`, and `append`.
`googleSheets`	`spreadsheetId`	string	Yes	Usually selected from the Drive picker.
`googleSheets`	`range` / `sheetName`	string	No	Supports values like `Sheet1!A1:Z100`.

Inputs Expected

excel read operations consume an uploaded spreadsheet file or cached fileIds.
excel write operations expect an array of objects or rows.
googleSheets reads do not need upstream input.
googleSheets writes/appends expect an array. Arrays of objects are converted into a header row plus values.

Variables Exported / Outputs

Node	Output shape	Example expressions
`excel` read	`Array<Record<string, any>>`	`{{excel_node.customer_name}}`, `{{excel_node.invoice_total}}`
`googleSheets` read	`Array<Record<string, string>>`	`{{sheets_node.email}}`, `{{sheets_node.status}}`
`excel` write	`Blob`	Usually consumed by download or cloud upload nodes.
`googleSheets` write/append	original array input	The node passes the written records forward.

Important implementation detail:

resolveVariables defaults array field access to the first row unless the workflow is inside a loop and context.inputData contains the current item.

Common Use Cases

Read a spreadsheet, loop over its rows, and fill a PDF or DOCX template.
Append processed metadata into a Google Sheet after a conversion step.
Split an Excel workbook by a column value and archive the resulting files.

Edge Cases

googleSheets is server-only. Client-side execution returns a failed status.
Native Google Sheets and Drive-hosted Excel files both work through the picker, but preview loading can fail if the selected item is not actually a spreadsheet.
excel write/merge/split output types differ from read output, so downstream nodes must match the chosen operation.