Token导航 LogoToken导航

Agent Skill · 研究检索

godot-adapt-3d-to-2d(godot 将 3d 调整为 2d)

godot-adapt-3d-to-2d 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

godot-adapt-3d-to-2d 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 38,052。

总安装 1,623周安装 69GitHub Stars 137下载 569核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:godot-adapt-3d-to-2d(godot 将 3d 调整为 2d)
来源仓库:https://github.com/thedivergentai/gd-agentic-skills
仓库路径:skills/godot-adapt-3d-to-2d
安装命令:
npx skills add https://github.com/thedivergentai/gd-agentic-skills --skill godot-adapt-3d-to-2d
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/thedivergentai/gd-agentic-skills --skill godot-adapt-3d-to-2d

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

godot-adapt-3d-to-2d 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。适用于信息检索和筛选任务。通过关键词、任务场景或来源线索快速定位候选结果。

安装命令:npx skills add https://github.com/thedivergentai/gd-agentic-skills --skill godot-adapt-3d-to-2d。建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

来源仓库内路径:skills/godot-adapt-3d-to-2d

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/thedivergentai/gd-agentic-skills --skill godot…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

godot-adapt-3d-to-2d 怎么安装?

在终端执行 npx skills add https://github.com/thedivergentai/gd-agentic-skills --skill godot-adapt-3d-…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

godot-adapt-3d-to-2d 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

godot-adapt-3d-to-2d 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

godot-adapt-3d-to-2d 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

godot-adapt-3d-to-2d 用的人多吗?

公开数据是累计安装 1,623 次,近一周安装 69 次,下载 569 次,GitHub 137 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

godot-adapt-3d-to-2d 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
godot-adapt-3d-to-2d

godot 将 3d 调整为 2d · 本页

1,623137只读未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Adapt: 3D to 2D

Expert guidance for simplifying 3D games into 2D (or 2.5D).

NEVER Do

  • NEVER remove Z-axis without gameplay compensation — Blindly flattening 3D to 2D removes spatial strategy. Add other depth mechanics (layers, jump height variations).
  • NEVER keep 3D collision shapes — Use simpler 2D shapes (CapsuleShape2D, RectangleShape2D). 3D shapes don't convert automatically.
  • NEVER use orthographic Camera3D as "2D mode" — Use actual Camera2D for proper 2D rendering pipeline and performance.
  • NEVER assume automatic performance gain — Poorly optimized 2D (too many draw calls, large sprite sheets) can be slower than optimized 3D.
  • NEVER forget to adjust gravity — 3D gravity is Vector3(0, -9.8, 0). 2D gravity is float (980 pixels/s²). Scale appropriately.

Available Scripts

MANDATORY: Read the appropriate script before implementing the corresponding pattern.

ortho_simulation.gd

Simulates 3D Z-axis height in 2D top-down games. Handles vertical velocity, gravity, sprite offset, and shadow scaling.

projection_utils.gd

Projects 3D world positions to 2D screen space for nameplates, healthbars, and targeting. Handles behind-camera detection and distance-based scaling.

isometric_math_core.gd

Expert utility generating translating between 2D Cartesian and True Isometric screenspace projection matrices without using 2D Node transforms.

depth_sorting_y_sort.gd

Expert dynamic Z-index Y-Sort script for fake 3D sorting isolated trees matching CanvasItem _update_sorting().

jump_z_axis_sim.gd

Complete CharacterBody2D snippet separating structural physical ground movement (X,Y) from a mathematically simulated jumping height (Z) in a topdown game.

parallax_depth_camera.gd

Fake Depth Camera applying varying offset algorithms to completely disparate CanvasLayers based on an index to simulate 3D camera translation panning.

hitbox_depth_manager.gd

Area2D derived class that requires explicit custom Z-height overlap (1D AABB collision) prior to validating 2D triggers to stop incorrect "ground vs air" collision in 2.5D.

fake_3d_shadows.gd

Sprite2D shadow simulator exploiting Godot 4.x Transform2D matrix skew shear to project and angle shadows away from a simulated 3D sun direction on a 2D floor.

billboard_sprite_manager.gd

8-directional FPS Doom-style sprite controller isolating the simulated 3D relative angle between a moving 2D CharacterBody and a Camera2D viewpoint.

nav_region_flattening.gd

Topdown 2D pathfinding workaround allowing "aerial" units to cross walls by leveraging multiple tiered 2D Navigation Layers instead of proper 3D verticality.

ortho_to_perspective_fx.gd

Screen space CanvasItem warp Shader simulating a Mode 7 / tabletop perspective pitch. Maps top screen coordinates via division pinching.

2d_lighting_normals.gd

Automatic programmatic generation of CanvasTexture combining base albedo and baked normal maps at runtime so Sprites correctly react to 2D PointLIGHTs like 3D geometry.


Why Go from 3D to 2D?

ReasonBenefit
Mobile performance5-10x faster on low-end devices
Simpler art pipelineSprites easier to create than 3D models
Faster iteration2D level design is quicker
AccessibilityLower hardware requirements
ClarityReduce visual clutter for puzzle/strategy games

Dimension Reduction Strategies

Strategy 1: True 2D (Remove Z-axis)

# Top-down or side-view
# Example: 3D isometric → 2D top-down

# Before (3D):
var velocity := Vector3(input.x, 0, input.y) * speed

# After (2D):
var velocity := Vector2(input.x, input.y) * speed

# Use case: Top-down shooters, RTS, turn-based strategy

Strategy 2: 2.5D (Fake depth with layers)

# Keep visual depth perception without Z-axis gameplay
# Use ParallaxBackground for depth layers

# Scene structure:
# ParallaxBackground
#   ├─ ParallaxLayer (far mountains, scroll slow)
#   ├─ ParallaxLayer (mid buildings, scroll medium)
#   └─ ParallaxLayer (near trees, scroll fast)

# player.gd
extends CharacterBody2D

func _ready() -> void:
    var parallax := get_node("../ParallaxBackground")
    parallax.scroll_base_scale = Vector2(0.5, 0.5)  # Parallax strength

Strategy 3: Fixed Perspective (Isometric Stay)

# Keep isometric/dimetric view but use 2D physics
# Use rotated sprites to simulate 3D angles

const ISO_ANGLE := deg_to_rad(-30)  # Isometric tilt

func world_to_iso(pos: Vector2) -> Vector2:
    return Vector2(
        pos.x - pos.y,
        (pos.x + pos.y) * 0.5
    )

func iso_to_world(iso_pos: Vector2) -> Vector2:
    return Vector2(
        (iso_pos.x + iso_pos.y * 2) * 0.5,
        (iso_pos.y * 2 - iso_pos.x) * 0.5
    )

Node Conversion

Physics Bodies

# CharacterBody3D → CharacterBody2D
extends CharacterBody3D  # Before

const SPEED = 5.0
const JUMP_VELOCITY = 4.5
const GRAVITY = 9.8

func _physics_process(delta: float) -> void:
    velocity.y -= GRAVITY * delta
    var input := Input.get_vector("left", "right", "forward", "back")
    velocity.x = input.x * SPEED
    velocity.z = input.y * SPEED
    move_and_slide()

# ⬇️ Convert to:

extends CharacterBody2D  # After

const SPEED = 300.0
const JUMP_VELOCITY = -400.0
const GRAVITY = 980.0  # Pixels per second squared

func _physics_process(delta: float) -> void:
    velocity.y += GRAVITY * delta
    var input := Input.get_vector("left", "right", "up", "down")
    velocity.x = input.x * SPEED
    # Note: No Z-axis. For platformer, use input.y for jump
    move_and_slide()

Camera Conversion

# Camera3D → Camera2D
# Before: Third-person 3D camera
extends SpringArm3D

@onready var camera: Camera3D = $Camera3D

func _process(delta: float) -> void:
    spring_length = 10.0
    rotate_y(Input.get_axis("cam_left", "cam_right") * delta)

# ⬇️ Convert to:

extends Camera2D  # After

@onready var player: CharacterBody2D = $"../Player"

func _process(delta: float) -> void:
    global_position = player.global_position
    zoom = Vector2(2.0, 2.0)  # Adjust to taste

Art Pipeline: 3D Models → Sprites

Option 1: Render Sprites from 3D (Automation)

# Use Godot to render 3D model from fixed angles
# sprite_renderer.gd (tool script)
@tool
extends Node3D

@export var model_path: String = "res://models/character.glb"
@export var output_dir: String = "res://sprites/"
@export var angles: int = 8  # 8-directional sprites
@export var render: bool = false:
    set(value):
        if value:
            render_sprites()

func render_sprites() -> void:
    var model := load(model_path).instantiate()
    add_child(model)

    var camera := Camera3D.new()
    camera.position = Vector3(0, 2, 5)
    camera.look_at(Vector3.ZERO)
    add_child(camera)

    var viewport := SubViewport.new()
    viewport.size = Vector2i(256, 256)
    viewport.transparent_bg = true
    viewport.add_child(camera)
    add_child(viewport)

    for i in range(angles):
        model.rotation.y = (TAU / angles) * i

        await RenderingServer.frame_post_draw
        var img := viewport.get_texture().get_image()
        img.save_png("%s/sprite_%d.png" % [output_dir, i])

    model.queue_free()
    camera.queue_free()
    viewport.queue_free()

Option 2: Manual Export (Blender)

# Blender Python script (run in Blender)
import bpy
import math

angles = 8
output_dir = "/path/to/sprites/"
model = bpy.data.objects["Character"]

for i in range(angles):
    model.rotation_euler.z = (2 * math.pi / angles) * i
    bpy.ops.render.render(write_still=True)
    bpy.data.images['Render Result'].save_render(
        filepath=f"{output_dir}/sprite_{i}.png"
    )

Option 3: Use Sprite3D as Reference

# Keep 3D model in editor, export  frame-by-frame

Physics Adjustments

Gravity Scaling

# 3D gravity (m/s²): 9.8
# 2D gravity (pixels/s²): Scale to pixel units

# If 1 meter = 100 pixels:
const GRAVITY_2D = 9.8 * 100  # = 980 pixels/s²

# Adjust jump velocity proportionally:
# 3D jump: 4.5 m/s
# 2D jump: -450 pixels/s

Collision Simplification

# 3D: CapsuleShape3D (16 segments, expensive)
var shape_3d := CapsuleShape3D.new()
shape_3d.radius = 0.5
shape_3d.height = 2.0

# 2D: CapsuleShape2D (much simpler)
var shape_2d := CapsuleShape2D.new()
shape_2d.radius = 16  # pixels
shape_2d.height = 64

Control Simplification

3D Free Movement → 2D Restricted

# 3D: Full 3D movement with camera-relative controls
var input_3d := Input.get_vector("left", "right", "forward", "back")
var camera_basis := camera.global_transform.basis
var direction := (camera_basis * Vector3(input_3d.x, 0, input_3d.y)).normalized()

# 2D: Simple 4-direction (or 8-direction with diagonals)
var input_2d := Input.get_vector("left", "right", "up", "down")
velocity = input_2d.normalized() * SPEED

Performance Gains

Expected Improvements

Metric3D2DImprovement
Draw calls100205x
GPU loadHighLow10x
Battery life (mobile)1 hour5 hours5x
RAM usage500MB100MB5x

Optimization Techniques

# 1. Use TileMapLayer instead of individual Sprite2D nodes
var tilemap := TileMapLayer.new()
tilemap.tile_set = load("res://tileset.tres")

# 2. Batch sprite rendering
# Use single large sprite sheet instead of individual textures

# 3. Reduce particle count
var godot-particles := GPUParticles2D.new()
godot-particles.amount = 50  # Down from 200 in 3D

UI Adaptation

# Most 3D games already use 2D UI (CanvasLayer)
# No changes needed!

# Just verify UI scaling for new aspect ratios
get_viewport().size_changed.connect(_on_viewport_resized)

func _on_viewport_resized() -> void:
    var viewport_size := get_viewport().get_visible_rect().size
    # Adjust UI anchors/margins

Edge Cases

Depth Sorting

# Problem: Overlapping sprites need sorting
# Solution: Use Y-sort or z_index

extends Sprite2D

func _ready() -> void:
    y_sort_enabled = true  # Auto-sort by Y position
    # Or set z_index manually:
    z_index = int(global_position.y)

Lost Spatial Audio

# 3D spatial audio (AudioStreamPlayer3D) → 2D panning (AudioStreamPlayer2D)

var audio_2d := AudioStreamPlayer2D.new()
audio_2d.stream = load("res://sounds/footstep.ogg")
audio_2d.max_distance = 1000.0  # 2D range
audio_2d.attenuation = 2.0
add_child(audio_2d)

Decision Tree: When to Simplify to 2D

FactorKeep 3DGo 2D
Target platformDesktop, consoleMobile, web
Art styleRealistic, immersiveStylized, retro
GameplayRequires 3D spaceWorks in 2D plane
PerformanceHave GPU budgetNeed 60 FPS on low-end
Team skills3D artists2D artists or pixel art

Reference

平台分布

Codex

34.53%
按下载量换算196

Claude

31.14%
按下载量换算177

Cursor

18.47%
按下载量换算105

Gemini CLI

9.09%
按下载量换算52

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills