Using Test Data

Test Data lets you store credentials and other inputs your tests depend on in a simple tabular format. Instead of rewriting prompts every time you need to swap a value, you define the data once and reuse it whenever you run a task.

It’s useful when you have values that repeat across tests or inputs you normally copy-paste during setup. With Test Data, you set it upfront and let the agent pull what it needs during execution.

How to Add Test Data

You’ll find Test Data in the left navigation panel, right after Validations and before Reports.

1. Click Add Test Data to create a new reusable data set.

2. Fill in the Basic Information:

   1. Name — This is a required field to give your test data a name.

   2. Description — optional context for your team.

   3. Slug — auto-generated in the format /your_slug.

3. Build your Data Table: Each column header in your table acts as a key. These are the labels the agent uses to understand what each value represents (for example: username, password, address_line_1, phone_number, etc). Each row represents a single, complete data entity you can pull during execution. Think of a row as one full set of values that belong together. Example: If you're storing different login credentials, your table might look like this:

Username	Password	Role
dev_user	pass123	admin
qa_user	qaPass!	tester

Here: username, password, and role are keys. Each row (dev_user..., qa_user...) is a reusable data entity. You can add as many rows and columns as you need.

1. Click Save Test Data. Your test data set is now ready to use.

Using Test Data During Execution

When you’re writing your prompt in the Execution window, type / to open the list of available test data sets. Select the dataset you want to use, and then choose the specific row you want the agent to pull values from. Each row represents one complete set of inputs, so the agent will use that row for the entire run. Once selected, the agent automatically fills in the values during execution. No extra setup or editing the prompt every time something changes.

We’re also working on a feature that will let you write a prompt once and run it across multiple rows (multiple datasets) in one go. This will make data-driven testing even simpler.

Editing or Updating Test Data

1. Open Test Data from the left navigation and click the dataset you want to update.

2. In the dataset view, make your changes: edit cell values, rename column headers, add new rows or columns, or remove ones you no longer need.

3. Review the table to confirm your edits and click on Save to make the changes.

4. The updated test data will be used for subsequent executions.