Skip to main content
Manage your AppSheet tables directly from Gumloop. Read, add, update, and delete rows in your AppSheet applications using simple, natural-language prompts to keep your data in sync across workflows.
Connect to any AppSheet app and perform CRUD operations on your tables without writing code or managing complex API calls.

How to Use MCP Nodes

What is AppSheet MCP?

AppSheet is Google’s no-code application development platform that lets you build apps from spreadsheets and databases. The AppSheet MCP node gives AI-powered access to your AppSheet applications, allowing you to interact with your app’s data tables using natural language. Describe what you need, and the node performs the action, returning structured data you can pass to other steps in your workflow.

What Can It Do for You?

  • Retrieve rows from any table in your AppSheet app with filters and conditions
  • Add new rows to keep your app data current as workflows run
  • Update existing rows to reflect changes from other systems or processes
  • Delete rows to clean up outdated or invalid records

Available Tools

ToolWhat It DoesExample Use
Get RowsRetrieve rows from an AppSheet table”Get all rows from table Customers in app app id where Status equals Active and return structured data with Name, Email, and Phone
Add RowsAdd new rows to an AppSheet table”Add a new row to table Orders in app app id with Customer=John Doe, Product=Widget, Quantity=5 and return the new row id”
Update RowsUpdate existing rows in an AppSheet table”Update rows in table Tasks in app app id where Assignee equals Jane to set Status=Complete and return updated row ids”
Delete RowsDelete rows from an AppSheet table”Delete rows from table Archive in app app id where Created Date is before 2024-01-01 and return deleted row count”

How to Use

1

Create Your AppSheet MCP Node

Go to your node library, search for AppSheet, and click “Create a node with AI”
2

Add Your Prompt

Drag the AppSheet MCP node to your canvas and add your prompt in the text box. Keep it focused on a single action and specify the app ID and table name.
3

Test Your Node

Run the node to see the results. If it works as expected, you’re all set. If you run into issues, check the troubleshooting tips below.
4

Save and Reuse

Once your AppSheet MCP node is working, save it to your library. You can now use this customized node in any workflow.

Example Prompts

Here are some prompts that work well with AppSheet MCP: Retrieve Active Records:
Get all rows from table `Inventory` in app `app id` where `In Stock` equals `true` and return structured data with `Product Name`, `SKU`, and `Quantity`
Add a New Entry:
Add a new row to table `Contacts` in app `app id` with `Name`=`Sarah Smith`, `Email`=`[email protected]`, `Company`=`Acme Corp` and return the new row id
Update Record Status:
Update rows in table `Projects` in app `app id` where `Project ID` equals `PRJ-123` to set `Status`=`In Progress`, `Updated By`=`System` and return updated row ids
Remove Old Records:
Delete rows from table `Logs` in app `app id` where `Timestamp` is before `2024-06-01` and return deleted row count
Always include the app ID and table name in your prompts to help the node locate the correct data. Use specific field names that match your AppSheet table schema for best results.

Troubleshooting

If your AppSheet MCP node isn’t working as expected, try these best practices:

Keep Prompts Simple and Specific

  • Good: “Get all rows from table Orders in app app id where Status equals Pending and return Order ID and Customer Name
  • Needs improvement: “Get all orders, update their status, send notifications, and archive the old ones”
For best results, create separate nodes for each action. AppSheet MCP works best with focused, single-action prompts.

Match What AppSheet Can Do

  • Good: “Add a new row to table Employees with Name=John, Department=Sales and return the new row id”
  • Needs improvement: “Create a new employee record and automatically send them a welcome email”
AppSheet MCP excels at table operations. For sending emails or notifications, connect to Gmail Sender or other notification nodes in your workflow.

Break Complex Tasks Into Steps

Instead of trying to do everything in one prompt (which might cause timeouts or errors):
Get all overdue tasks, update their priority to High, notify the assignees, and generate a summary report
Break this into smaller, focused nodes that each handle one task:
1

Step 1: Get Overdue Tasks

Get all rows from table Tasks in app app id where Due Date is before today and Status equals Open and return Task ID, Title, and Assignee
2

Step 2: Update Priority

Update rows in table Tasks in app app id where Task ID is in task ids from step 1 to set Priority=High and return updated row ids
3

Step 3: Notify and Report

Use Ask AI and notification nodes to process the task list and alert assignees
Connect these nodes sequentially in your workflow. The task IDs from Step 1 feed into Step 2, and the task details feed into Step 3 for notifications.

Focus on Data Operations

AppSheet MCP is great at reading and writing data to your AppSheet tables. For analysis or content creation, connect it to other nodes. Example:
  • Good prompt: “Get all rows from table Sales in app app id where Region equals West and return Amount, Date, and Product
  • Needs improvement: “Get all sales data and create a quarterly revenue analysis with charts”
Use AI for analysis or summarization. Keep AppSheet MCP focused on retrieving and updating table data, then pass the structured data to Ask AI for reports.

Troubleshooting Node Creation

If you’re seeing empty outputs in the node creation window (or if you’ve already created the node, hover over it and click “Edit”), use the chat interface to prompt the AI to add debug logs and verify the API response. Specifically mention that you received empty outputs.
In the node creation window (or if you’ve already created the node, hover over it and click “Edit”), use the chat interface to describe in detail what you expected versus what you received.
MCP node creation often requires a few tweaks. Use the chat interface in the node creation window to refine filters, output fields, or pagination. The AI will adjust the node based on your feedback.

Need More Help?