Bevy genre router for Puzzle. Keywords: Bevy, Puzzle, genre, ECS, router.
npx skills add https://github.com/thedivergentai/Game-Design-Agentic-Skills --skill bevy-genre-puzzle
Genre router. Domain work lives in the Bevy skills below.
Load this for grid rules and an undo stack you own.
Update or the fixed timestep from bevy-master. Order them with .chain() when a later system must see the result.[dependencies]
bevy = "0.20"
Each tile stores a grid coordinate and whether it is a goal. The undo stack is a resource of prior coordinates. Read input, push history, then slide. Those systems share Tile, so they chain.
#[derive(Clone, Copy)]
struct GridPos {
x: i32,
y: i32,
}
#[derive(Component)]
struct Tile {
pos: GridPos,
goal: bool,
}
#[derive(Component)]
struct PlayerToken;
#[derive(Resource, Default)]
struct UndoStack {
steps: Vec<GridPos>,
}
fn boot(mut commands: Commands) {
commands.insert_resource(UndoStack { steps: Vec::new() });
commands.spawn((
PlayerToken,
Tile {
pos: GridPos { x: 0, y: 0 },
goal: false,
},
));
commands.spawn(Tile {
pos: GridPos { x: 2, y: 0 },
goal: true,
});
}
fn read_move(keys: Res<ButtonInput<KeyCode>>) -> GridPos {
if keys.just_pressed(KeyCode::ArrowRight) {
GridPos { x: 1, y: 0 }
} else if keys.just_pressed(KeyCode::ArrowLeft) {
GridPos { x: -1, y: 0 }
} else if keys.just_pressed(KeyCode::ArrowUp) {
GridPos { x: 0, y: 1 }
} else if keys.just_pressed(KeyCode::ArrowDown) {
GridPos { x: 0, y: -1 }
} else {
GridPos { x: 0, y: 0 }
}
}
fn push_undo(
keys: Res<ButtonInput<KeyCode>>,
players: Query<&Tile, With<PlayerToken>>,
mut undo: ResMut<UndoStack>,
) {
if keys.just_pressed(KeyCode::KeyZ) {
return;
}
let delta = read_move(keys);
if delta.x == 0 && delta.y == 0 {
return;
}
if let Some(tile) = players.iter().next() {
undo.steps.push(tile.pos);
}
}
fn slide_player(
keys: Res<ButtonInput<KeyCode>>,
mut players: Query<&mut Tile, With<PlayerToken>>,
) {
if keys.just_pressed(KeyCode::KeyZ) {
return;
}
let delta = read_move(keys);
for mut tile in &mut players {
tile.pos.x += delta.x;
tile.pos.y += delta.y;
}
}
fn undo_move(
keys: Res<ButtonInput<KeyCode>>,
mut undo: ResMut<UndoStack>,
mut players: Query<&mut Tile, With<PlayerToken>>,
) {
if !keys.just_pressed(KeyCode::KeyZ) {
return;
}
let Some(prev) = undo.steps.pop() else {
return;
};
for mut tile in &mut players {
tile.pos = prev;
}
}
fn install(app: &mut App) {
app.add_systems(Startup, boot).add_systems(
Update,
(undo_move, push_undo, slide_player).chain(),
);
}
The puzzle is solved when the player token shares a cell with a goal tile.
#[derive(Clone, Copy, PartialEq, Eq)]
struct GridPos {
x: i32,
y: i32,
}
#[derive(Component)]
struct Tile {
pos: GridPos,
goal: bool,
}
#[derive(Component)]
struct PlayerToken;
fn solved(players: Query<&Tile, With<PlayerToken>>, goals: Query<&Tile>) -> bool {
let Some(player) = players.iter().next() else {
return false;
};
goals.iter().any(|tile| tile.goal && tile.pos == player.pos)
}
push_undo and slide_player both call read_move, which reads ButtonInput twice in one frame. just_pressed stays true for the frame, so both see the same key.
| Piece | Bevy idiom |
| --- | --- |
| Genre content | Entities plus the domain skills linked below |
| Run-wide director | A Resource |
| Loop notices | bevy-events-messaging |
GridPos. UndoStack keeps prior player coordinates. Bevy has no puzzle-grid type; the example above is the data.Target: Bevy 0.20+.
Core ECS types: Component, Resource, Query, Commands, Startup, Update, Fixed / FixedMain (bevy::time), .chain(), Plugin, DefaultPlugins, MinimalPlugins. Messages and observers are an ownership slot. See bevy-events-messaging for their APIs. Systems run in parallel unless .chain() orders the ones that conflict.
| Piece | Owner | Who may mutate | Schedule |
| --- | --- | --- | --- |
| Data on one entity | Component | One system via Query | Update or FixedMain |
| One value for the whole app | Resource | One system | Startup to insert, then the schedule that reads it |
| Spawn | Commands | The spawning system | Startup for the initial world; later schedules for runtime spawn |
| Order that must not race | .chain() | The chained systems only | Same schedule |
| Plugin boundary | Plugin | build registers systems and resources | Before the app runs |
Borrow rule: two systems that both mutably query the same component must be ordered (.chain()) or split so only one writes. Unrelated systems stay parallel. That is the ECS page rule, not a guess at a system-set API.
| If you need | Do this | Do not |
| --- | --- | --- |
| Per-entity data | Component + Query | One script that owns the whole tree |
| App-wide config or clock | Resource | Copy the value onto every entity |
| Stable simulation step | FixedMain and the Fixed clock | Tie simulation to raw frame Update |
| Startup spawn | Commands in Startup | Spawn from a render system |
| Cross-system notice that is not shared state | Message or observer slot in bevy-events-messaging | A new event type invented in this file |
| An API outside the documented set | Check the official docs first | A made-up crate or API |
Load references/ecs-ownership.md for the ownership table.
/0.20.0/ in docs URLs. WHY: links stay on latest; the 0.20 target is stated once in prose..chain() orders them.Official pages stay on unversioned permalinks. The library pin on bevy-master stays 0.20.
latest branch.Component, systems, Query, Startup, Update, and .chain().Plugin, DefaultPlugins, MinimalPlugins, and add_plugins.Take thedivergentai/bevy-genre-puzzle from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.