What this prompt does
This prompt makes the AI a senior healthcare product designer specifying a medication tracking screen tightly enough to build, with components and states named. It defines a today schedule grouped by time slot, one-tap taken/skipped logging with undo, an adherence week view, inline interaction warnings when two scheduled meds conflict, refill-due alerts with doses remaining, and accessibility built for the audience. The [audience] placeholder shapes the accessibility and tone, and [a11y] sets the floor the components must meet.
The structure works because adherence UIs live or die on accessibility, and the people who most need them are often older or low-vision. The prompt insists on large touch targets, high contrast, and screen-reader labels on every status, and it makes logging a single tap with an undo affordance, since extra taps quietly kill adherence. The [stack] variable keeps the component tree mapped to your framework, with each dose card showing name, dose, and status. Grouping the schedule by time slot means the screen mirrors how people actually take medication through the day, morning, midday, evening, rather than as one long flat list, which makes it far easier to see at a glance what is due now and what has already been handled.
When to use it
- Building a medication reminder or adherence app
- Designing for an
[audience]such as older adults managing several daily meds - You need an
[a11y]floor like WCAG AA with large text and high contrast - Adding one-tap logging with undo for taken and skipped doses
- Surfacing drug interaction warnings inline between scheduled meds
- Showing refill-due alerts with a count of doses remaining
Example output
You get a component tree with named props and the key states for each component. The today schedule is grouped by time slot with dose cards showing name, dose, and status; logging is one tap with an undo affordance and a confirmed state; the adherence week view contrasts taken vs missed with an empty state; interaction warnings appear inline; and refill alerts show doses remaining with a reorder CTA. Every status carries a screen-reader label.
Pro tips
- Set
[audience]precisely; designing for older adults changes touch-target sizes and contrast more than for general users - Use
[a11y]to state your real floor so the model bakes contrast and large text into every component - Keep logging a single tap with undo; if the output adds a second tap, ask it to simplify, since extra taps reduce adherence
- Match
[stack]to your framework so the component tree is buildable - Ask it to confirm a screen-reader label on every status, not just the visible ones
- If interaction warnings feel buried, request they surface inline at the conflicting dose rather than in a separate alerts area
