Validations

This guide walks you through adding validation rules to your sheets — automatic checks that flag data quality issues like missing values, bad emails, or numbers outside an expected range.

Add Validation Rules with the Agent

The quickest way: ask the agent.

Add validation rules to flag empty emails and invalid phone numbers

The agent applies the rules and every row in your sheet is checked immediately. Cells that fail show color-coded flags — you can hover over any flag to see exactly what's wrong.

You can keep refining in plain language.

Also flag any order amounts under 0 or over 10,000

The agent adds a range check to the column. No configuration screens to hunt through.

Add Validation Rules Manually

Validations are rules attached to a sheet, targeting one field (column) and checking every row against a condition you define. While the agent is the quickest way, you can also manually configure rules on individual fields via the sheet operations panel.

What Validated Data Looks Like

When a cell fails a validation rule, it gets a colored highlight and a small indicator icon. Hover over the cell to see the validation message.

The color tells you the severity at a glance:

• Red — error: Data is incorrect or incomplete.
• Yellow — warning: Data looks suspicious but might be intentional.
• Blue — info: An informational flag, not necessarily a problem.

Validation Types

Each rule type checks for something specific. Here's every type available and what it catches.

Required

Flags any row where the field is empty.

Flag rows missing a company name

Email

Checks whether the value is a valid email address.

Add an email validation to the contact email column

Phone

Checks whether the value looks like a phone number.

Validate phone numbers in the mobile column

URL

Checks whether the value is a properly formatted web address.

Flag invalid URLs in the website field

Pattern

Checks whether the value matches a specific format you define using a pattern (regular expression).

Flag any SKU that doesn't follow the format ABC-1234

Range

Checks whether a number falls within a minimum, maximum, or both.

Flag any price below 0 or above 5000

Integer

Checks whether a number is a whole number — no decimals.

Flag quantities that aren't whole numbers

Unique

Flags duplicate values in a field.

Make sure there are no duplicate email addresses

Custom

Write your own condition when the built-in types don't cover what you need.

Flag any row where the status is "inactive" but the balance is greater than zero

Severity Levels

Every validation rule has a severity, indicating the level of urgency.

Severity	Color	When to use
Error	Red	Data is incorrect or incomplete. Needs action.
Warning	Yellow	Data looks off but might be valid. Worth a review.
Info	Blue	A flag for visibility. Nothing necessarily wrong.

If you don't specify a severity, rules default to error.

Add a warning-level validation that flags orders over $50,000

Removing Validation Rules

Validation rules can be removed by asking the agent, which will clear the rule and its associated flags immediately without affecting your data.

Remove the phone number validation from this sheet

Next Steps

• Formulas — create calculated fields that feed into your validation rules
• Importing Data — bring in data and let your validation rules catch issues on arrival
• Field Types & Configuration — set up columns with the right types so validations work accurately