The Frame-First AI Cheatsheet Download PDF, English Download PDF, Nederlands

The Frame-First AI Cheatsheet

You don't need the full solution first. You need a clear frame: context, goals, and constraints. Then leverage AI to iterate so you can quickly test and validate ideas.

00 · Start in conversation

Before you touch any tool or code, work through the frame in dialogue with AI: bring your research, insights and competitor examples, and ask AI to find what is missing. Refine all three sections together.

01

Context

Who, what, why

The situation the prototype is for.

Who
The user, stakeholder, team, or role it is for.
What
What they want to do, decide, understand, or improve.
Why & pain
The friction: what feels slow, risky, blocked, or unclear.
Current reality
What happens today, and the workaround that already exists.
02

Goals

What should improve

The outcome you want the user to reach.

User goals
The goal, job, or decision it should help them reach.
Success
What it should help you learn, prove, compare, or clarify.
Design Style
The emotion it should create in use, and the brand direction to match. Calm, playful, premium, gameful.
Interaction
Low-friction, guided, scannable, fast, confidence-building.
03

Constraints

What must stay true

Where limits become design boundaries.

Security & privacy
Data or systems that can never be touched or exposed.
Technical limits
No backend, no database, internal only, offline only.
Fidelity
Wireframe, clickable flow, or working shell with fake data.
Scope limits
What is intentionally out of scope this round.

Blockers turned to Constraints

Every blocker hides a rule. Drop the "so we can't", and rewrite what is left as "It must…" or "It cannot…".

"AI can't use our internal data."
→
It must use synthetic or redacted data.
"We don't have backend access yet."
→
It must run on the front-end with mocked responses.
"It's too risky to build fully."
→
It cannot go past a clickable prototype this round.
04

Build, validate, iterate

The frame is step one, not the whole job. Build the first version fast, put it in front of real users, and act on what they do. Move in small loops: change one piece, or rethink the whole idea if it isn't landing. Then run it again.