html-game is a small, fast HTML game that runs in the browser. The product intent is a lightweight, instantly playable arcade/reflex experience that loads quickly, runs smoothly, and gives a casual anonymous visitor something delightful to poke within seconds. The specific game type was left open by the user ("just random one"), so the builder has chosen a simple arcade/reflex-style browser game — a pickup/collect arcade game titled PICKUP — where the player moves a chunky coral capsule around a cream playfield to collect mint targets while avoiding hazards.
The audience is a single active human role: the Player — a casual, anonymous visitor who arrives with zero context, wants to be playing within about two seconds, and measures success in "one more go" rather than in comprehension. There is no account, no setup, no onboarding flow, and no marketing surface. The game is the page.
html-game is delivered as a single-page browser experience built from HTML, CSS, and JavaScript. It has two cohesive destinations:
Both destinations are owned by the application and require no identity, no login, and no persistence. The Player is anonymous throughout. There is no backend, no database, no provider-owned surface, and no external destination required for current scope.
Actors:
Accepted behavior: a small, fast, browser-playable arcade game with a landing overlay that dissolves into an already-running playfield, responsive physics-driven play, immediate score feedback, a game-over state, and restart.
Narrow exclusions: no accounts, no login, no persistence, no leaderboards, no multiplayer, no separate "how to play" page, no multi-step onboarding, no loading spinners or asset preloaders before play, no photographic or stock imagery, no marketing hero/feature grid/footer.
The product is a self-contained, first-party browser game. Delivery is headless of any server: the entire experience is client-side HTML/CSS/JavaScript, so there is no provider-owned surface, no external destination, and no backend responsibility in current scope. Access is anonymous and public — the Player never establishes identity, and no durable actor-specific state is created or resumed, so application-owned identity is neither required nor inferred.
The current horizon covers the Landing and Game destinations and the play lifecycle they support. Future ideas (for example, persistent best-score storage across sessions, additional game modes, or shareable results) are explicitly out of current scope and are listed only in the future section; they must not appear in current pages or acceptance criteria.
Not applicable — no reference directive in this project declares content_source.
FR-1 — Small, fast HTML game (explicit) As a Player, I should be able to open a small, fast HTML game in my browser and be playing within about two seconds, so that I get immediate lightweight play with no setup.
FR-2 — Randomly chosen arcade/reflex game (explicit) As a Player, I should be able to play a simple arcade/reflex-style browser game (the specific type left to the builder's choice), so that I get a delightful, momentum-driven toy rather than a product to evaluate.
FR-3 — Start a session from the Landing overlay (required_inference) As a Player, I should be able to press a single PLAY capsule on the Landing overlay to start a session, so that the CTA never navigates and there is zero load-to-play gap.
FR-4 — Responsive physics-driven play (required_inference) As a Player, I should be able to move the player object around the playfield with responsive, physics-driven motion, so that every interaction has a springy, physical answer.
FR-5 — Collect targets with immediate feedback (required_inference) As a Player, I should be able to collect mint targets and see immediate physical feedback, so that the score readout lives in the world and the feedback is physical.
FR-6 — Miss feedback and recovery (required_inference) As a Player, I should be able to see a short screen-shake and a desaturating flash when I miss, so that misses are legible and the round continues.
FR-7 — Game-over / end-of-round state and restart (required_inference) As a Player, I should be able to see a game-over/end-of-round state with my final score and restart with a single capsule press, so that I can go again immediately.
FR-8 — Mute control (required_inference) As a Player, I should be able to toggle mute from the HUD rail, so that I can control the game's audio without leaving the playfield.
html-game with zero context. They did not search for a specific game; they were handed a small, fast HTML game and want to be playing within about two seconds. They are on a browser, likely on a desktop or laptop, and they will leave immediately if anything asks them to set up, sign in, or wait.What makes the Player's work different from a generic "user": there is no account, no setup, no comprehension task, and no evaluation. The Player's entire relationship with the product is tactile — they poke the toy, learn the rules by touching things, and measure success in "one more go."
html-game URL in a browser.The creative direction is authoritative for this section. The muse is Bruno Simon; the headline is "Playable toy-box physics" — a tiny, instant-load browser arcade game whose first screen is already the toy.
Colour tokens (light mode):
| Role | Hex | Usage |
|---|---|---|
| Background | #FFF6E4 | Warm pastel-cream ground; replaces white |
| Surface | #FFFFFF | Overlay card, floating label card |
| Text | #20213A | Type and the chunky 3px outlines that hold every shape together |
| Primary | #FF6B4A | Player/CTA colour; carries the loudest area of the hero |
| Accent | #2FD6A5 | Target/hazard and score-positive states |
| Muted | #8E8AA6 | Secondary labels and disabled states only |
Colour blocking, not gradients: each zone of the playfield is a flat field with a hard edge and a soft drop shadow.
Typography:
Shape language: Toy-box geometry — low-poly forms with 2–3 visible facets, capsules and rounded blocks with 20–28px radii, every element wrapped in a 3px ink outline, and soft offset shadows (0 6px 0 rgba(32,33,58,.18)) that make pieces look like they can be picked up. Buttons are capsules that physically depress 4px on press, not fade on hover.
Layout: The game IS the page — a full-viewport playfield panel with a thin HUD rail across the top (score left, best right, mute far right) and a single floating label card in the lower-left naming the game. No marketing hero, no feature grid, no footer. The landing state is the same viewport with a translucent overlay card holding the title, one sentence of rules, and a single PLAY capsule — pressing it dissolves the card and the world is already running underneath.
Imagery: No photography, no stock art. Everything is drawn in code: flat low-poly shapes with two-tone facet shading, a soft circular ground shadow under each object, and a few procedural confetti particles on score events. The "art" is the geometry and the colour blocking itself.
Avoid: centred headline + subtext + button stacked in the middle of the viewport; any blue/indigo primary or gradient-blob background; Inter, Roboto, Arial, Poppins, or system-ui for any text; a grid of identical hover-lift feature cards; loading spinners or asset preloaders before play can begin; photographic or stock imagery of any kind; soft blurred shadows and glassmorphism panels; a separate "how to play" page or multi-step onboarding. The generic indigo/blue-on-white SaaS template is forbidden for this project.
The hero is the game, not a picture of the game.
The public entry (Landing) is a full-bleed cream playfield with a hard-edged coral ground band along the bottom third and a mint target cluster floating above it. The player object is a chunky coral capsule with an ink outline and a visible contact shadow, roughly 1/12 of the viewport width, positioned left-of-centre so the composition reads as "something is about to happen moving right." The title PICKUP is set at 72px Fredoka with an ink outline and coral offset shadow, overlapping the top-left corner of the playfield and partially cropped by the viewport edge — a poster gesture, not a centred SaaS headline. A single capsule PLAY button is cut into the coral ground band as a physical hole rather than sitting on top of it. The scene is already animating behind the overlay, so the hero is the game, not a picture of the game.
Signature moves:
Interaction Model: Animated Motion Tempo: expressive Hero Dimensionality: layered_2d
Landing Hero Motion Brief
NFR-1 — Small and fast (explicit) The game must be small and fast: lightweight, quick to load, and quick to run. Rationale: explicit hard constraint in the authoritative user evidence. Acceptance: the playfield renders and animates immediately on arrival; no spinner, preloader, or onboarding step appears before play.
NFR-2 — Browser-playable HTML delivery (explicit) The game must be delivered as HTML playable in a browser. Rationale: explicit hard constraint in the authoritative user evidence. Acceptance: the game runs in a standard browser with no install step.
NFR-3 — Responsive input latency (required_inference) Player input must produce visible motion within the same frame or the next, so that the game feels responsive. Rationale: indispensable to the accepted "responsive play" and "immediate feedback" behavior. Acceptance: the player object responds to input without perceptible lag.
NFR-4 — Readability on the cream ground (required_inference) Body text must never fall below 16px, and HUD numerals must use tabular figures at 32px so the score does not jitter. Rationale: indispensable to the accepted visual direction and to legibility on the warm pastel-cream ground. Acceptance: all body text is at least 16px; the score readout does not shift horizontally as digits change.
NFR-5 — No pre-play loading (explicit) No loading spinners or asset preloaders may appear before play can begin. Rationale: explicit exclusion in the creative direction. Acceptance: the playfield is animating on first paint.
Assumptions:
Constraints:
Future (not in current scope):
#20213A outline wrapped around every shape.0 6px 0 rgba(32,33,58,.18) shadow that makes pieces look like they can be picked up.No completed page designs yet.
Completed design pages will appear here when they are ready to preview.
No completed page designs yet.
Completed design pages will appear here when they are ready to preview.
No comments yet. Be the first!