Before you begin
Open the existing project so your instruction is grounded in the current app.
Write the instruction from the user’s perspective, then state the exact result you want to inspect.
Write the instruction
The best prompts do not sound technical. They make five decisions clear.
1. User
Who is using the app?
2. Job
What are they trying to finish?
3. Input
What information do they provide?
4. Rules
What should the app do with that information?
5. Result
What should appear, and what does a good result look like?
Use this structure
“Build an app for [user] to [job].
The user provides a defined input.
The app should [rule one], [rule two], and [rule three].
Return the result as [format].
A good result means [success criterion].
Do not [important constraint].”
You do not need to name a model, framework, database, or AI technique unless one is genuinely
required. Describe the work as clearly as the person doing it would describe it.
That gives Empromptu AI enough information to make useful technical decisions without forcing
you to become the technical specification.
Common mistakes
Omitting the user, input, or output.
Using goals such as “modern,” “smart,” or “better” without observable criteria.
Combining the first workflow with later features.
Frequently asked questions
How long should a prompt be?
Long enough to define the user, job, input, rules, output, and important constraints. Clear structure
matters more than word count.
Can I make several changes in one prompt?
Small related changes can be combined, but changes to different systems are easier to test and
recover when requested separately.