📄 README for sheets2ai

sheets2ai is a Google Apps Script utility that converts structured data from Google Sheets into a clean, readable TXT format optimized for Large Language Models (LLMs). The tool processes every non-empty cell in the selected spreadsheet and saves the information into a text file on Google Drive.

While LLMs can understand natural language, they often struggle to interpret complex, formula-laden spreadsheet data. sheets2ai solves this by exporting spreadsheet data — including formulas, values, and cell types — into a simple textual format that is both human- and machine-readable.

• 🔍 Simplifies spreadsheet structures for better LLM interpretation.
• 📂 Organizes data consistently across multiple sheets.
• 📄 Includes cell formulas and data types for context.
• 🛠️ User-configurable, allowing exclusion of irrelevant sheets.
• 📤 Saves directly to Google Drive, ready for LLM processing.

This script runs entirely in the Google Apps Script environment, requiring no external installations.

• Google Apps Script
  Built-in scripting platform for automating tasks in Google Workspace.
  Why Needed? Used to read Google Sheets, process cell data, and save TXT files in Google Drive.

• SpreadsheetApp: Accesses and reads spreadsheet data.
• DriveApp: Writes TXT files to Google Drive.
• Logger: Logs execution progress and errors.

💡 No downloads or installations are required beyond access to a Google account and Google Sheets.

1. Open the Google Sheet you want to export.
2. Click on Extensions > Apps Script.

1. Replace any existing code in the Apps Script editor with the full script provided in this repository.
2. Click File > Save and give your project a name.

At the top of the script, locate and modify these lines:

const spreadsheetId = 'YOUR_SPREADSHEET_ID_HERE'; // Replace with actual Sheet ID
const outputFolderId = 'YOUR_OUTPUT_FOLDER_ID_HERE'; // Replace with Drive folder ID
const excludedSheetNames = []; // Optional: Add sheet names to exclude

• How to find spreadsheetId:

  • It's the string in your spreadsheet URL between /d/ and /edit.
  • Example: https://docs.google.com/spreadsheets/d/123ABC456DEF789GHI/edit → ID = 123ABC456DEF789GHI

• How to find outputFolderId:

  • Open your target Google Drive folder, and copy the string after /folders/ in the URL.

1. In the Apps Script editor, click the function dropdown (next to the debug icon).
2. Select exportSheetDataToTxt.
3. Click the ▶️ Run button.
4. Authorize the script when prompted.

• Loops through all sheets (excluding those you specify).
• For every non-empty cell, extracts:
  • Sheet name
  • Cell address (e.g., Sheet1!A2)
  • Cell value
  • Formula (if any)
  • Data type (text, number, integer, boolean, date, NaN)

• Excluded Sheets: Modify excludedSheetNames to ignore certain sheets:

const excludedSheetNames = ["Instructions", "Calculations"];

• A .txt file saved to your Google Drive folder, containing:

Sheet: Sheet1
  Cell: Sheet1!A1, Value: Sales, Formula: data, Data Type: text
  Cell: Sheet1!B2, Value: 1234.56, Formula: data, Data Type: number

Scenario: You want to feed tabular business data into GPT-4 for analysis.
How It Helps: sheets2ai exports structured data with context (formulas, types), making it LLM-friendly.
Example Output:

Cell: Sheet1!C3, Value: =SUM(A1:A2), Formula: =SUM(A1:A2), Data Type: number

Scenario: You need to archive your financial model logic for an audit.
How It Helps: Formulas and values are exported clearly.
Example Output:

Cell: Sheet2!E10, Value: 2025-01-01, Formula: data, Data Type: date

Scenario: You’re building a model that learns from spreadsheet structures.
How It Helps: Clean and consistent output of cell-level info enables easy data ingestion.
Example Output:

Cell: Data!F5, Value: TRUE, Formula: data, Data Type: boolean

• This script and its functionality may change at any time without notice.
• Such changes may render parts of this README obsolete.
• The provided code is delivered "as-is" without any guarantees, warranties, or assurances of functionality, reliability, or compatibility.
• No commitment is made to maintain or update this code or its documentation.