Przejdź do treści
Biblioteka skilli
Projektowanie i UX · Planowanie

ui-motion

/ui-motion

Decyduje, czy coś powinno się animować i jak: czas, sprężyny, choreografię, gesty, redukcję ruchu i budżet klatek w aplikacjach oraz na stronach.

Licencja
MIT
Ostatnia zmiana repozytorium
08.11.2026

Ocena i bezpieczeństwo

90/ 100Klasa: dobry
Format25 / 25
Frontmatter z nazwą i opisem, poprawna nazwa, opis do 1024 znaków, nazwa zgodna z folderem.
Odkrywalność25 / 25
Opis mówi, kiedy użyć skilla, ma sensowną długość i jest w 3. osobie.
Treść25 / 35
Co najmniej 150 słów i do 500 linii, nagłówki, listy kroków, przykłady albo kod.
Higiena15 / 15
Bez zaślepek typu TODO, bez ścieżek Windows, z licencją.
Ryzykobrak sygnałów w skanie
Pliki w paczce
2
Skrypty
0
Komentarze HTML
0
Adresy URL w SKILL.md
0
Przestarzałość

Brak sygnałów. Repozytorium zmienione 08.11.2026, bez odwołań do wycofanych modeli i API.

Licencja

MIT, otwarta: można używać i zmieniać, z zachowaniem informacji o autorze.

Co robi i kiedy

Decyduje, czy coś powinno się animować i jak: czas, sprężyny, choreografię, gesty, redukcję ruchu i budżet klatek w aplikacjach oraz na stronach.

Wyzwalacze z opisu

floatycheapslowjanky
Oryginalny opis (EN)

Interface animation and interaction craft — deciding whether something should animate at all, then how: timing, springs, choreography, gestures, reduced motion and frame-budget performance. Use when building or reviewing app/website motion, popovers, sheets, drag interactions, page transitions, hover and press feedback, or when an interface feels "floaty", "cheap", "slow" or "janky". Apple-grade fluidity on a web frame budget.

Jak działa

  1. Should it animate at all?
  2. The timing table
  3. The four rules that prevent "floaty"
  4. Origin and relationship
  5. What to animate (and what never to)
  6. Gestures
  7. Reduced motion is not optional
  8. The frame budget
  9. Reviewing motion (yours or someone else's)
Sekcje z nagłówków SKILL.md, w oryginale · 125 linii, 994 słów
Podgląd SKILL.md
---
name: ui-motion
description: >-
  Interface animation and interaction craft — deciding whether something should animate at all,
  then how: timing, springs, choreography, gestures, reduced motion and frame-budget performance.
  Use when building or reviewing app/website motion, popovers, sheets, drag interactions, page
  transitions, hover and press feedback, or when an interface feels "floaty", "cheap", "slow" or
  "janky". Apple-grade fluidity on a web frame budget.
license: MIT
---

# UI motion

An interface that moves well feels physical and inevitable. One that moves badly
feels like a slideshow with opinions. The difference is almost never the library.

## Should it animate at all?

Ask in this order. The first "no" means do not animate it.

1. **Did something change that the user did not do?** → animate it in (they need to
   know it appeared, and where from).
2. **Did the user just do it?** → animate only if the change is instant-but-continuous
   (a drag, a knob, a scroll). Continuity, not decoration.
3. **Is it on a high-frequency surface?** (a key repeat, a table sort, a keystroke,
   a hover trail) → **no.** Frequency and motion are inversely proportional: the
   more often it happens, the less it should move.
4. **Does it communicate a relationship?** (this panel came from that button; this
   row moved to that list) → yes, and the motion must show the relationship.
5. **Is it purely decorative?** → no, unless it is the one moment the piece is
   allowed and it costs no frames.

## The timing table

| Interaction | Duration | Notes |
|---|---|---|
| press / tap feedback | 100–170 ms | must feel instantaneous; transform only |
| hover | 130–270 ms | travel ≤ 2 px; never animate box-shadow |
| tooltip | 270–400 ms | origin-aware, no delay on show if it followed a click |
| popover / dropdown | 270–400 ms | scale from the trigger corner, 0.96 → 1 |
| sheet / modal | 470–800 ms | one spring; backdrop fades 1.4× slower than the panel |
| page / route | 200–400 ms | content in, chrome stays put |
| list reorder | 200–350 ms | items move along the shortest path |
| expand / collapse | 250–400 ms | animate `transform` + `scaleY`, never `height` |

Above 800 ms is a deliberate, watched moment. Below 100 ms nothing is perceived
as motion — either use no transition, or make it 130 ms and feel intentional.

## The four rules that prevent "floaty"

1. **One spring, at most, per view.** Springs are for physical things the user
   manipulated. A spring on a text label reads as a toy.
2. **Motion is proportional to distance.** A 4 px move is not 400 ms. Above 40 px of
   travel, add no more than 100 ms per additional 100 px.
3. **Everything that enters exits the same way.** Same curve, same duration,
   reversed. Consistency is what reads as craft.
4. **The end pose is exact.** No half-eased remainder, no 0.998 scale. Set the
   final state explicitly.

## Origin and relationship

The most legible motion in an interface is *where things came from*:

- A popover scales from the corner nearest its trigger, not from its centre.
- A detail view expands from the card that was tapped (a shared-element
  transition), not by sliding in from the right edge.
- A moved row animates along the path it actually took.
- A deleted item shrinks toward the trash, or toward where it went — not straight
  down and out.

If you cannot state the relationship, the animation is decoration. Delete it.

## What to animate (and what never to)

**Animate:** `transform` (translate, scale, rotate), `opacity`, `filter` (sparingly —
blur is expensive), `clip-path` (for reveals).

**Never animate:** `width`, `height`, `top`, `left`, `margin`, `padding`, `border-width`,
`box-shadow`, `background-position`, or anything that forces layout or paint on
every frame. These are the reason interfaces feel janky on real hardware.

Never `transition: all`. It is the same thing as "animate whatever happens", which
is what "cheap" means.

## Gestures

- **Track the finger 1:1** while dragging. Any easing during the drag feels broken.
- **On release, decide in one frame:** commit if past threshold or fast enough
  (velocity matters more than distance), else spring back.
- **Interruptible.** A second gesture during an animation takes over from the
  current visual position, never from the animation's start.
- **Rubber-band past the boundary** by ~12% of the travel, then resist.
- **Haptic/visual acknowledgement on the frame the gesture is recognised**, not
  when the animation completes.

## Reduced motion is not optional

Respect `prefers-reduced-motion: reduce`:

- Remove all decorative motion, parallax, and auto-playing movement.
- Keep state changes, but make them instant, or use a 100 ms opacity change.
- **No information may be conveyed by motion alone.** If the user cannot see the
  animation, they must still understand the outcome.
- Never hide content behind an intersection-observer reveal that also fails for
  reduced motion — the classic bug where the page looks empty.

## The frame budget

At 60 fps you have 16.7 ms per frame. Real hardware (a mid-range Android, a
thermal-throttled laptop) has less. So:

- Composite-only properties keep the work off the main thread.
- Blur, large shadows, and 3+ simultaneous transforms will drop frames.
- Animate at most 2–3 elements simultaneously; stagger the rest.
- Test with 6× CPU throttling. If it holds, it ships.

Details, spring maths and a performance audit procedure: `references/craft.md`.

## Reviewing motion (yours or someone else's)

Do not say "feels off". Say which of these it is:

| Symptom | Likely cause | Fix |
|---|---|---|
| floaty | duration too long, or easing too soft | ×0.7 duration, `expo.out` |
| cheap | spring on everything, or `transition: all` | one spring max, explicit properties |
| janky | animating layout/paint properties | transform/opacity only |
| confusing | origin or relationship unstated | animate from the trigger |
| slow | over 400 ms on a frequent interaction | halve it |
| distracting | high-frequency surface, or no state change | remove it |
| broken | end pose inexact, or not interruptible | set final state, take over on gesture |
| inaccessible | motion is the only signal | add a static state change |

Default to flagging. Approval must be earned, not inferred.
Pliki w paczce (2)
  • SKILL.md 6 KB
  • references/craft.md 5 KB

Wersje

RepozytoriumWynikLicencja
Ta strona · jamescai-md/motionframe-edit90MIT
whoamnesia/astudio-motion90MIT

Migawka bazy z 13.11.2026 · wersja paczki 85f86796f617 (skrót SHA-256 całej paczki)

Pobrania w ostatnich 30 dniach: 0 · ranking

Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.

Podobne skille

  • 100motion
    Projektuje, buduje i sprawdza animacje webowe: przejścia, wejścia i wyjścia, gesty, sprężyny i ruch zredukowany.
  • 100kk-motion
    Doradza w animacji interfejsów: czy ruch jest potrzebny, czas i easing, sprężyny, hover, gesty, przerwania i wydajność, w oparciu o szkołę Birmana, Apple i Emila Kowalskiego.
  • 100motion
    Rozstrzyga, czy element powinien się animować, i buduje lub przegląda ruch: cel, właściwości, easing, czas trwania, sprężyny, przerwania i reduced motion.
  • 100motion-craft
    Doradza, czy i jak animować w sieci: czas trwania, krzywe, sprężyna czy przejście, prefers-reduced-motion, a także wejścia, hover, szuflady, odsłanianie przy przewijaniu i gesty przeciągania.
  • 100ux-motion
    Projektuje, wdraża i sprawdza ruch w UI: przejścia, gesty i informację zwrotną, a także naprawia wolne lub rwące animacje.
  • 100swift-motion-animate
    Dodaje animacje w aplikacji SwiftUI: od decyzji, czy w ogóle animować, przez sprężyny i przerwania, po zmniejszenie ruchu i haptykę.

Uwagi i zgłoszenia

Wiadomość trafi do Bartosza Sokolińskiego. Adres e-mail jest potrzebny tylko do odpowiedzi. Szczegóły w polityce prywatności.