AI prompts for form design (DNN)

The AI Designer responds best to prompts that name fields, types and rules explicitly. These patterns are tested wording — paste, adapt, send.

Creating a form (dashboard → ✨ Create with AI)

Create an event registration form: full name, work email (validated), company, ticket type as radio buttons (Standard / VIP / Student), attendance date, dietary notes as long text, and a consent checkbox that must be ticked.

Build a job application form with a two-column name row, email, phone, a file upload for the CV (PDF only, max 5 MB), years of experience as a number, and a rating for "How well do you know .NET?" from 1 to 5.

Editing the open form (builder → ✨ AI Designer)

Add a star rating field named "Overall satisfaction" after the last field.

Make Full Name required and limit Short Text to 200 characters.

Put First name and Last name side by side in one row.

Add a dropdown "Department" with options HR, IT, Finance, Operations — default to IT.

Split this form into two steps: contact details first, everything else on step 2.

Conditional logic

Show the "Company name" field only when "Are you registering for a company?" is Yes.

If Amount is greater than 5000, require the "Manager email" field.

Database-aware design (the panel's Database tab)

Create a form for the Stores table: dropdowns for Country and Currency from their tables, and insert each submission into dbo.Stores.

Add a dropdown "Vendor" whose options come from dbo.Vendors ordered by name.

What makes prompts work

• Name the field and its type — "a star rating field named X", not "some way to rate".
• One intent per message for edits; batch whole structures only at create-time.
• State the rule, not the implementation — "must be a valid email" beats "add a regex".
• The AI applies validated operations and the form stays a Draft — experiment freely, undo is right there, and nothing is public until you publish.