ux playground

playground 101

4 of 4

Work through a project

Six steps, each shown through one project: the appointment card. Read each one when your project reaches it.

Understand the problem

You approve the brief

Before anything is built, the agent asks about the problem and hands you a brief. Here's one for the appointment card.

Brief for your approval · appointment card

Problem
Some patients arrive not ready for their scan, so the visit runs late or is rebooked. Their prep instructions came in an email weeks earlier.Names what goes wrong, not a screen to build.
For
Patients in the last two days before their scan, including someone a reminder call doesn't reach, someone travelling with only a phone, and someone anxious about the scanner.The situations that make it hard, not a list of people.
Success
The day before, each patient sees what they still need to do for their scan, on the device they use. Nobody calls the clinic to ask.Something you can check.
Can't change
The confirmation email. Prep content comes from what clinics already send.
You reply"Approve.""Add people who don't read English well."Nothing is built until you approve.
Not:

"We need a prep checklist card on the dashboard home."

That's one answer, not the problem.

The approved brief becomes user needs in the project's README: outcomes for the patient, each with a status the hub card counts. Working from a linked PRD? The agent flags each user need the PRD doesn't decide yet, so you can take it back to the PM.

User needs · appointment card

  • ✓ DoneAs a patient, I can see the day before what I still need to do for my scan.
  • △ PartialAs a patient on my phone, I can tick off a step without signing in again.
  • ☐ GapAs a patient who reads English as a second language, I can understand each step.
  • ✦ NewAs a patient travelling to my scan, I see times in my clinic's time zone.

Done, Partial and Gap are agreed. New is raised, not agreed yet.

When you get here, try:

Help me turn this into a problem, not a solution: "Add a progress bar to
booking." Ask me who it's for and what goes wrong for them today, one question
at a time. Then write the Problem, For and Success lines, and the situations
that would make it hardest to get right.

Set a direction

You pick the direction

Taste is easier to recognise than to describe. Show the agent what you mean, then react to what it makes.

Taste note · appointment card

Feel
Calm and certain, like a boarding pass: when, where, and what to bring.Says how it should feel, in your words.
Borrow
An airline boarding pass: one big time, the date beside it. A clinic reminder text: plain, two lines.What each reference gives, not just a link.
Avoid
A countdown or a cheerful checklist. Prep shouldn't feel like a test.A line not to cross.
Builds on
The playground taste. This note adds only what's new for this card.Shared taste still holds. You clarify when there's an explicit departure from Prenuvo patterns.
You reply"Like the boarding pass, but quieter.""Debate this."React in a sentence. "Debate this" has agents argue two directions.
Not:

"Make it clean and modern."

A guess at what clean means to you.

Share whatever shows the feel: a site or app you like, screenshots, Mobbin screens, a napkin sketch.

When you get here, try:

Here are three references for my project: [links or screenshots]. Tell me
what they have in common, what to borrow and what to avoid. Then show me two
directions that use them, side by side, so I can react.

Explore concepts

You steer as it builds

Describing the perfect component up front is hard. Reacting to ten real options is easy.

Component board · appointment card

  • 1 · Kit card today
  • 2 · Time first
  • 3 · Date block
  • 4 · Map thumbnail
  • 5 · Prep list inline
  • 6 · Time, then prep linkConcept A
  • 7 · Countdown
  • 8 · TimelineConcept B
  • 9 · Compact row
  • 10 · Ticket stub
Starts from
The kit's own card, option 1.What production shows today, so you compare against what's real.
States
Each option shown upcoming, today and rescheduled, for two synthetic patients.Real states and data, never lorem ipsum.
Then
Choose one option, mix and match two, or build two as concepts to compare live.A concept is a version of your project you switch between in the prototype's toolbar.
Decision log
Build 6 and 8 as concepts A and B. "People look for when before what. Not sure yet if the timeline calms or crowds."Your choice and your words are kept, so the next person knows why.
Not:

"Make the appointment card better."

One guess, one round at a time.

When you get here, try:

Make a component board for the appointment card on the dashboard kit: ten
options side by side, starting from the kit's own card. Show each one for an
upcoming scan, a scan today and a rescheduled one, using two different
synthetic patients. I'll react, then choose which to build as concepts and tell you why.

Check your work

You pick which findings to fix

A crit walks your concept as the people who'll use it, through four lenses: usability, interaction, brand and accessibility. It finds who you missed while it's still cheap to fix, keeps the platform consistent, and shows which choices you made and which the agent guessed. Here's one finding from a crit of concept A of the appointment card.

Finding · appointment card, concept A

Major
Skye is travelling. The card says "8:00 AM" in his phone's time zone, not the clinic's, so he plans to arrive an hour late.Interaction lens. Decide before handoff.
Check it
Switch to Skye Stopover and set the browser to another time zone.You can see it for yourself.
Against
The brief: each patient sees what they need for their scan, on the device they use.Held to something agreed, not taste.
Fix
Show the clinic's time and name its zone, "8:00 AM PT", the way booking already does.It follows the platform pattern, so every surface agrees.
Works
The prep link sits one tap from the time.A crit says what works, too.
You reply"Fix this one.""Add it as a user need."Nothing changes until you pick.
Not:

"The card feels confusing."

No patient, no way to check it, nothing agreed. That's an opinion.

Run a crit whenever the concept changes what someone sees, then just on what you fixed. Each one ends by offering to tidy small things on the same page.

When you get here, try:

Walk the synthetic patients through this project and critique it. Give me the
three findings that matter most. For each, say which patients it affects, what
they'd experience, which platform pattern applies if there is one, and the fix
you'd suggest. Then stop and let me choose.

Share and test

You choose what feedback you need from real people

Push your work at the end of every session, then share it on purpose.

End of every session

Safe
Saved on GitHub, not just your laptop.
Visible
The project card shows open work, with a live link.
Current
Small, frequent pushes merge cleanly. Weeks of local work don't.
You say"Push my work."Unfinished is fine: it merges early, and your next push picks up from there.
Not:

"I'll push it when it's ready."

Nobody sees it, and it's gone if your laptop is.

To share on purpose, each audience gets its own kind of link:

Who should see itSayWhat they get
Your team"Push my work."A live copy builds from the PR on its own, and the project card shows it as open work.
Prenuvo reviewers without GitHub"Get concept A ready for the internal hub."A link behind Google sign-in, with comments if you want them (the publish asks), and the card shows In testing. A hub owner does the publish.
Research participants"Get concepts A to D ready for the external hub."A password-protected link for a study session. A hub owner does the publish.

Every link is pinned to one commit, so later changes never move a link someone is already using.

When you get here, try:

Push my work. Then tell me what each audience would see
if I published concept A to the internal hub, and what I'd need before testing it with
participants on the external hub.

Hand it off

You decide when it's ready

We used to draw every state and leave engineering to work out the logic between them. A handoff now says how the concept works. Here's the start of one for concept A of the appointment card.

Handoff · appointment card, concept A

Source
Concept A is the source. If a Figma frame disagrees, the prototype was updated last.The working prototype, not a frame.
Rules
Show the time first, in the clinic's time zone, and name the zone. Show the prep link from two days before the scan until check-in. Don't port the synthetic patients' names or dates.What to build, kept apart from demo data.
Situations
Upcoming · today · rescheduled · prep all done · no prep needed · patient in another time zone · loading · appointment can't load.Every case the one card has to handle.
Not:

"Here are all 24 states in Figma."

Engineering has to guess the rules between them.

A handoff also names the components to use and the questions engineering still has to answer, then lists every state and the exact copy. Larger features get use cases and a spec too.

When you get here, try:

Hand off concept A of my project. Check the kit is up to date, show me exactly
what the concept changes from production, and ask me before you create the
ticket.