Explain the limits of the AI and generator

Worked example

A procedural bug report might say seed 17, build alpha2, start at entrance, move right, expected clear corridor, actual blocking wall. The seed makes reproduction practical. Scaffold C-GRID — A deterministic grid with a guaranteed clear route Setup: Attach to Node2D GridLab. This example draws, not physics, so the assignment adds matching wall collision only after validation. Grid is 8 columns by 6 rows at 40 pixels per cell. Row 0 and column 7 are a guaranteed clear route from (0,0) to (7,5). RNG seed is a nonpersonal integer. ```gdscript extends Node2D const COLS: int = 8 const ROWS: int = 6 const CELL: int = 40 var cells: Array = [] func _ready() -> void: var rng := RandomNumberGenerator.new() rng.seed = 2468 for y in range(ROWS): var row: Array[int] = [] for x in range(COLS): var blocked: bool = y != 0 and x != COLS - 1 and rng.randf() < 0.25 row.append(1 if blocked else 0) cells.append(row) queue_redraw() func _draw() -> void: for y in range(cells.size()): for x in range(cells[y].size()): var color := Color.DIM_GRAY if cells[y][x] == 1 else Color.BEIGE draw_rect(Rect2(x * CELL, y * CELL, CELL - 2, CELL - 2), color) ``` Scaffold C-GUARD-AI — A readable patrol/chase/return state machine Setup: Create CharacterBody2D Guard with visible circle or polygon and matching collider. Layer 4 mask 2 for walls. Attach script and assign target to your Player instance in Inspector. Level is one open rectangle; this is steering, not navigation around walls. Damage remains a separate sensor/manager event. Draw detection radii on a design diagram before tuning values. ```gdscript extends CharacterBody2D enum Mode { PATROL, CHASE, RETURN } @export var target: Node2D @export var speed: float = 70.0 var mode: Mode = Mode.PATROL var home: Vector2 var patrol_target: Vector2 func _ready() -> void: home = global_position patrol_target = home + Vector2(80, 0) func _physics_process(delta: float) -> void: if not is_instance_valid(target): velocity = Vector2.ZERO return var distance: float = global_position.distance_to(target.global_position) if mode != Mode.CHASE and distance < 120.0: mode = Mode.CHASE elif mode == Mode.CHASE and distance > 180.0: mode = Mode.RETURN var destination: Vector2 if mode == Mode.CHASE: destination = target.global_position elif mode == Mode.RETURN: destination = home if global_position.distance_to(home) <= maxf(3.0, speed * delta): mode = Mode.PATROL else: destination = patrol_target if global_position.distance_to(patrol_target) <= maxf(3.0, speed * delta): patrol_target = home if patrol_target != home else home + Vector2(80, 0) var offset := destination - global_position velocity = offset.normalized() * minf(speed, offset.length() / maxf(delta, 0.001)) move_and_slide() ```

A deterministic grid with a guaranteed clear route

GDScript (Godot 4, not Scratch, Python, JavaScript or C#). Scaffold C-GRID. Copy starts from this text. The desk does not run it.

Scene and node setup

Attach to Node2D GridLab. This example draws, not physics, so the assignment adds matching wall collision only after validation. Grid is 8 columns by 6 rows at 40 pixels per cell. Row 0 and column 7 are a guaranteed clear route from (0,0) to (7,5). RNG seed is a nonpersonal integer.

extends Node2D
const COLS: int = 8
const ROWS: int = 6
const CELL: int = 40
var cells: Array = []
func _ready() -> void:
    var rng := RandomNumberGenerator.new()
    rng.seed = 2468
    for y in range(ROWS):
        var row: Array[int] = []
        for x in range(COLS):
            var blocked: bool = y != 0 and x != COLS - 1 and rng.randf() < 0.25
            row.append(1 if blocked else 0)
        cells.append(row)
    queue_redraw()
func _draw() -> void:
    for y in range(cells.size()):
        for x in range(cells[y].size()):
            var color := Color.DIM_GRAY if cells[y][x] == 1 else Color.BEIGE
            draw_rect(Rect2(x * CELL, y * CELL, CELL - 2, CELL - 2), color)

Checks to run in the editor

• Same seed and version produce same grid.
• All cells on top row and rightmost column are empty.

Native Godot 4.7.2.stable.official.ed1daf0bf headless compile and representative node-tree/runtime tests passed; see native-qa-results.json. Visual UI, audible output, exports and website submission still need separate checks.

A readable patrol/chase/return state machine

GDScript (Godot 4, not Scratch, Python, JavaScript or C#). Scaffold C-GUARD-AI. Copy starts from this text. The desk does not run it.

Scene and node setup

Create CharacterBody2D Guard with visible circle or polygon and matching collider. Layer 4 mask 2 for walls. Attach script and assign target to your Player instance in Inspector. Level is one open rectangle; this is steering, not navigation around walls. Damage remains a separate sensor/manager event. Draw detection radii on a design diagram before tuning values.

extends CharacterBody2D
enum Mode { PATROL, CHASE, RETURN }
@export var target: Node2D
@export var speed: float = 70.0
var mode: Mode = Mode.PATROL
var home: Vector2
var patrol_target: Vector2
func _ready() -> void:
    home = global_position
    patrol_target = home + Vector2(80, 0)
func _physics_process(delta: float) -> void:
    if not is_instance_valid(target):
        velocity = Vector2.ZERO
        return
    var distance: float = global_position.distance_to(target.global_position)
    if mode != Mode.CHASE and distance < 120.0:
        mode = Mode.CHASE
    elif mode == Mode.CHASE and distance > 180.0:
        mode = Mode.RETURN
    var destination: Vector2
    if mode == Mode.CHASE:
        destination = target.global_position
    elif mode == Mode.RETURN:
        destination = home
        if global_position.distance_to(home) <= maxf(3.0, speed * delta):
            mode = Mode.PATROL
    else:
        destination = patrol_target
        if global_position.distance_to(patrol_target) <= maxf(3.0, speed * delta):
            patrol_target = home if patrol_target != home else home + Vector2(80, 0)
    var offset := destination - global_position
    velocity = offset.normalized() * minf(speed, offset.length() / maxf(delta, 0.001))
    move_and_slide()

Checks to run in the editor

• Far player: guard patrols 80 pixels from home.
• Within 120 pixels: chase.
• Beyond 180 pixels: return, then patrol.
• Missing target stops safely rather than throwing an invalid-instance error.

Native Godot 4.7.2.stable.official.ed1daf0bf headless compile and representative node-tree/runtime tests passed; see native-qa-results.json. Visual UI, audible output, exports and website submission still need separate checks.

Explain the limits of the AI and generator — evidence task

1. 1. Write a 200-word note or private narrated diagram with code references.
2. 2. Include two supported capabilities and two limitations.

Deliverable. AI/generation explanation and supporting evidence.

Acceptance checks

• VGC-L134-T1. Claims match the actual implementation.
• VGC-L134-T2. At least one limitation is demonstrated by a test or diagram.

Understanding question. What does the reserved route prove, and what does it not prove?

Submit readable text or photos/scans of each numbered design page. Audio explanation may supplement a diagram or text, not replace requested artifacts.

If this is still shaky. Keep guard patrol in a reserved open lane and the player route guaranteed. Avoid adding maze pathfinding to this project. Start with the first assigned task and re-run its acceptance test before adding the remaining feature.

Optional extra. Add one optional guard route only if the generator can prove its lane remains clear. Document one changed parameter, your prediction, and the observed effect; optional, not required for completion.

Turn in a typed answer, a JPEG, PNG, or WebP photo, a PDF scan, or an MP3, WAV, M4A, MP4, or WebM file up to 4 MB. A parent can log paper. One https address is stored as text and is not opened or run. Zip and code files are refused.

• Your first 2D game

  Free public tutorial; no account; external asset download is optional, not used by this course.

  · Opening this link does not submit the lesson.