Skip to main content

Google Sheets Integration

Connecting Google Sheets lets ScaleFlow and AI agents read, write, and manage spreadsheet data you have granted access to. Agents can read ranges, append rows, update cells, or create new spreadsheets through MCP tools.

Use this integration when you want to:

  • Save leads, tickets, or conversation notes to a shared spreadsheet.
  • Let AI agents read pricing tables, inventory, or FAQ data stored in sheets.
  • Automate periodic reporting (read a range -> summarize -> write results).

Before You Start​

Prepare the following:

  • A Google account (Gmail or Google Workspace) with access to the spreadsheets you need.
  • Permission to manage integrations in ScaleFlow (you can see the Integrations menu and click Connect).
  • A spreadsheet already created in Google Sheets (or let the agent create a new one through MCP).
  • About 5-10 minutes for the first connection.

File access: ScaleFlow uses the Google Drive file scope — it can only access spreadsheets you select through Google Picker when configuring an agent, or files created by the integration. Spreadsheets that were not selected or shared will return permission errors.

Step Overview​

StepWhereWhat to do
1ScaleFlowIntegrations -> Google Sheets -> Connect
2GoogleSign in and Allow read/write access to sheets
3ScaleFlowConfirm the connection is Connected and run Test connection
4ScaleFlow(Optional) Open the Display tab and rename the connection
5ScaleFlowAttach the connection to an AI Agent and select spreadsheets
6ScaleFlowWrite Instructions and run Test Version

Step 1 - Connect Google Sheets in ScaleFlow​

  1. In ScaleFlow, open the left menu -> Integrations.

  2. Find the Google Sheets card under Third-party Integrations -> click Manage Integration.

    Google Sheets integration card

  3. On the setup screen, review Permissions required:

    • Read spreadsheet rows — read data for workflows and agents.
    • Append new rows — add rows from automation.

    Google Sheets setup permissions

  4. Click Continue with Google.

  5. In the Google window, select the Google account used for your spreadsheets.

  6. Click Allow to grant ScaleFlow the requested permissions.

  7. After authentication, you are redirected to ScaleFlow. The connection appears with Connected status.

    Google Sheets connections list

Step 2 - Manage the Connection​

Open Integrations -> Google Sheets and select a connection.

Display tab​

  • Connection name — change the display name (for example: Google Sheets — Sales, Google Sheets — Ops).
  • Google account metadata (read-only):
    • Account name
    • Google email
    • Connected on

Google Sheets Display tab

AI and Automation tab​

This tab shows:

  • An overview of spreadsheet workflows with AI.
  • Google Sheets actions — the list of MCP tools.
  • Recommended setup flows — lead capture, AI reporting, human review queue.

Google Sheets AI and Automation tab

Quick actions​

ActionWhen to use it
Test connectionAfter connecting, or when you suspect the token expired
ReconnectGoogle revoked access or you need to sign in again
DisconnectTemporarily stop the connection
DeleteRemove the connection from ScaleFlow

Google Sheets connection actions

Step 3 - Attach Google Sheets to an AI Agent​

  1. Open AI -> Agents and select an agent (or create a new one).

  2. In the right sidebar, open the Capabilities tab.

  3. Under Integrations, add the Google Sheets connection.

  4. Click Select spreadsheets and choose one or more spreadsheets through Google Picker.

    Agent select Google Sheets spreadsheets

  5. Click Save Draft -> Test Version.

    Test Google Sheets

Important: The agent can only work with spreadsheets selected in step 4 (or spreadsheets the agent creates through MCP). If you skip file selection, many tools will return No spreadsheet ID provided or configured.

Step 4 - Write Instructions to Test the Agent​

Paste the following into the Runbook / Instructions field (keep your existing tone/format and add this block):

#GoogleSheets: You can access Google Sheets through MCP tools.

## Rules
- When the user asks for spreadsheet data -> use google_sheets__read_rows with an A1 notation range (for example Sheet1!A1:D20).
- When the user provides information to save -> use google_sheets__append_row with sheetName and values.
- Only use google_sheets__update_values, google_sheets__clear_values, and google_sheets__delete_sheet when the user explicitly requests it.
- Do not invent data — always read from the sheet before answering questions about numbers.

## Default sheet for testing
- spreadsheetId: (fill in the ID from the Google Sheets URL, or let the agent use the spreadsheet selected in Capabilities)
- sheetName: Sheet1
- test read range: Sheet1!A1:C10

Suggested test prompts:

GoalTest question
Read data"Read the first 10 rows in Sheet1 and summarize them for me."
Append row"Add a new row: Name=John A, Email=[email protected], Source=Chat."
Metadata"What tabs (sheets) does this spreadsheet have?"

MCP Tools - What Can an Agent Do?​

Google Sheets MCP actions

Read Rows (google_sheets__read_rows)​

Reads data from an A1 notation range.

ParameterRequiredDescription
rangeYesFor example: Sheet1!A1:D20, Leads!A:E
spreadsheetIdNoSpreadsheet ID; leave empty if already selected in agent Capabilities

Append Row (google_sheets__append_row)​

Adds a new row at the bottom of a sheet.

ParameterRequiredDescription
sheetNameYesTab name, for example Sheet1, Leads
valuesYesArray of column values, for example ["John A", "[email protected]", "Chat"]
spreadsheetIdNoAs above

Update Values (google_sheets__update_values)​

Overwrites values in a range.

ParameterRequiredDescription
rangeYesA1 notation range
valuesYes2D array, for example [["A1", "B1"], ["A2", "B2"]]
spreadsheetIdNoAs above

Clear Values (google_sheets__clear_values)​

Clears content in a range (keeps the sheet structure).

ParameterRequiredDescription
rangeYesA1 notation range
spreadsheetIdNoAs above

Get Spreadsheet (google_sheets__get_spreadsheet)​

Returns metadata: file name, tab list, sheet IDs.

ParameterRequiredDescription
spreadsheetIdNo**Required if not configured in the agent

Create Spreadsheet (google_sheets__create_spreadsheet)​

Creates a new spreadsheet (ScaleFlow has access because the app created it).

ParameterRequiredDescription
titleYesNew file name

Add Sheet (google_sheets__add_sheet)​

Adds a new tab to an existing spreadsheet.

ParameterRequiredDescription
titleYesNew tab name
spreadsheetIdNoAs above

Delete Sheet (google_sheets__delete_sheet)​

Deletes a tab from a spreadsheet.

ParameterRequiredDescription
sheetIdYesNumeric tab ID (from Get Spreadsheet)
spreadsheetIdNoAs above

Get the Spreadsheet ID from a Google URL​

Open the spreadsheet in your browser. The URL looks like:

https://docs.google.com/spreadsheets/d/1BxiMVs0XRA5nFMdKvBdBZjgmUUqptlbs74OgvE2upms/edit

The part between /d/ and /edit is the spreadsheetId:

1BxiMVs0XRA5nFMdKvBdBZjgmUUqptlbs74OgvE2upms

In most cases, you do not need to copy this manually — selecting spreadsheets through Select spreadsheets in agent Capabilities is enough.

1. Lead capture to sheet​

Customer chats -> agent collects name, email, and source -> Append Row to the Leads tab.

2. AI reporting workspace​

Agent Read Rows from a KPI table -> summarizes for the user -> Update Values in the end-of-month report cell.

3. Human review queue​

Agent prepares a data row -> writes it to a review sheet -> the team approves directly in Google Sheets.

Real-World Workflow​

  1. An admin connects Google Sheets with the Sales team's Google account.
  2. The admin creates a Lead Capture agent and selects the Leads 2026 spreadsheet in Capabilities.
  3. The admin writes Instructions: append a row whenever a new lead appears.
  4. The admin runs Test Version -> Publish -> attaches the agent to Smart Assistant.
  5. When Google revokes access, the admin Reconnects and runs Test connection.

Quick Troubleshooting​

Error: No spreadsheet ID provided or configured​

  • In the agent -> Capabilities -> select at least one spreadsheet through Select spreadsheets.
  • Or pass spreadsheetId explicitly in Instructions or the tool call.

Error: Insufficient permissions / 403​

  • The spreadsheet was not selected through Google Picker in the agent.
  • Or the file belongs to a different Google account — reconnect with the correct account.
  • Try Reconnect on the integration and select the spreadsheet again.

Error: Spreadsheet not found / 404​

  • Confirm the spreadsheetId is correct.
  • The file may have been deleted or the user lost access.

Error: Google OAuth did not return a refresh token​

Agent does not call Sheets tools​

  • Confirm the Capabilities tab includes the Google Sheets connection.
  • Instructions must clearly state when to use read/append tools.
  • Click Save Draft before running Test Version.

Integrations menu is missing​

  • Ask your administrator for permission to manage integrations.

Best Practices​

  • Use clear connection names by team or purpose (Google Sheets — Marketing).
  • Use dedicated tabs (Leads, Report) instead of overwriting the default Sheet1.
  • Enable write/delete tools (update, clear, delete sheet) only when truly needed.
  • Test on a spreadsheet copy before using a production file.
  • Reconnect periodically if the team changes Google Workspace accounts.