Godot Genre Puzzle logo

Godot Genre Puzzle

Community
thedivergentai
godot-genre-puzzle

Expert blueprint for puzzle games including undo systems (Command pattern for state reversal), grid-based logic (Sokoban-style mechanics), non-verbal tutorials (teach through level design), win condition checking, state management, and visual feedback (instant confirmation of valid moves). Use for logic puzzles, physics puzzles, or match-3 games. Trigger keywords: puzzle_game, undo_system, command_pattern, grid_logic, non_verbal_tutorial, state_management.

Overview

Publisherthedivergentai
RepositoryGD-Agentic-Skills
Skill namegodot-genre-puzzle
Stars
727
Forks
43
Bundled files
17
LicenseLGPL-3.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 17 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by thedivergentai on GitHub. Read the source before you install it.

Installation

Install the Godot Genre Puzzle AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/thedivergentai/GD-Agentic-Skills.git /tmp/GD-Agentic-Skills
mkdir -p .claude/skills
cp -r /tmp/GD-Agentic-Skills/skills/godot-genre-puzzle .claude/skills/godot-genre-puzzle
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Godot Genre Puzzle in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Godot Genre Puzzle on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Godot Genre Puzzle is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

NEVER Do (Expert Anti-Patterns)

Design & Player Experience

  • NEVER punish experimentation; strictly provide Undo/Reset functionality to allow risk-free hypothesis testing.
  • NEVER require pixel-perfect input for logic puzzles; strictly use Grid Snapping or large, forgiving hitboxes.
  • NEVER allow undetected Soft-Locks (unsolvable states); strictly notify the player or provide immediate backtracking.
  • NEVER hide the rules of the world; strictly ensure visual feedback is instant and unambiguous (e.g., powered wires must glow).
  • NEVER skip the Non-Verbal Tutorial phase; strictly introduce mechanics in isolation before combining them.

Grid Logic & State

  • NEVER use floating-point numbers (Vector2) for grid coordinates; strictly use Vector2i to prevent precision drift.
  • NEVER use _process() for grid-state or win-condition validation; strictly trigger checks only when a piece moves.
  • NEVER rely on the SceneTree structure as the source of truth; strictly maintain grid data in a separate script/dictionary.
  • NEVER modify a Dictionary or Array size while iterating over it; strictly use a copy or a separate queue for modifications.
  • NEVER calculate heavy recursive solvers in _process(); strictly cache results or use threaded workers for solve-checks.
  • NEVER ignore diagonal rules in pathfinding; strictly configure AStarGrid2D.diagonal_mode correctly.

Architecture & Performance

  • NEVER ship dual undo authorities; strictly use Godot's built-in UndoRedo via puzzle_undo_manager.gd — do not also maintain a hand-rolled Command stack.
  • NEVER intermingle "do" and "undo" logic in the same function; strictly maintain separation for predictable rollbacks.
  • NEVER use exact floating-point equality (==); strictly use is_equal_approx() for spatial constraints.
  • NEVER use load() for resetting large rooms dynamically; strictly use ResourceLoader.load_threaded_request().
  • NEVER leave Tween objects unreferenced; strictly kill active tweens before starting new movement on the same object.

🛠 Expert Components (scripts/)

MANDATORY reads before implementing the matching system:

  1. puzzle_undo_manager.gd — sole undo authority (UndoRedo)
  2. grid_manager.gd — Vector2i grid as truth
  3. puzzle_state_validator.gd — soft-lock / win checks on commit

Original Expert Patterns

Modular Components

Do NOT load command_undo_redo.gd — legacy hand-rolled Command stack superseded by UndoRedo.


Core Loop

  1. Observe → 2. Hypothesize → 3. Commit move → 4. Validate → 5. Undo/Reset if needed

Decision Trees

Genre → scripts

Puzzle typeMANDATORY reads
Sokoban / push gridgrid_manager.gd, grid_tween_mover.gd, puzzle_undo_manager.gd, puzzle_state_validator.gd
Match-3match_three_logic.gd, grid_tween_mover.gd, puzzle_undo_manager.gd
Physics / spatialSnap on settle → then same undo/validator path; do not treat rigid bodies as truth
Hints / solverspuzzle_pathfinder.gd (AStarGrid2D.diagonal_mode)

Undo authority

NeedAction
Player undo/redoOnly puzzle_undo_manager.gd
Level editor historyStill UndoRedo — wrap editor actions the same way

Skill Chain

PhaseSkillsPurpose
1. Datadictionaries, resourcesGrid truth, level .tres
2. Inputgodot-input-handlingSnapped moves
3. Motiongodot-tweeningPiece Tweens
4. AI/hintsgodot-navigation-pathfindingA* hints
5. Persistgodot-save-load-systemsLevel / progress

Common Pitfalls

PitfallSolution
Dual undo APIsDelete Command-stack path; use UndoRedo manager only
Win check in _processValidate on move commit via state validator
Float grid coordsVector2i only

MANDATORY for depth beyond decision trees and script catalog: puzzle-elite-implementations.md. Do NOT Load on first-pass wiring — use bundled scripts/ first.

Architecture Overview

1. Command Pattern (Undo System)

Essential for puzzle games. Never punish testing.

gdscript

## Godot-Specific Tips

*   **Tweens**: Use `create_tween()` for all grid movements. It feels much better than instant snapping.
*   **Custom Resources**: Store level data (layout, starting positions) in `.tres` files for easy editing in the Inspector.
*   **Signals**: Use signals like `state_changed` to update UI/Visuals decoupled from the logic.

---

## Reference

> Progressive disclosure: open Official Documentation links only when researching a specific API;
> load Related Skills when routing work to a peer domain — do not preload the whole lattice.

### Official Documentation
- [UndoRedo](https://docs.godotengine.org/en/stable/classes/class_undoredo.html) — Built-in action history for do/undo/redo so puzzle experimentation does not require a hand-rolled command stack.
- [AStarGrid2D](https://docs.godotengine.org/en/stable/classes/class_astargrid2d.html) — Uniform-grid pathfinding (`diagonal_mode`, `jumping_enabled`) for hints and reachability on Sokoban-style boards.
- [Tween](https://docs.godotengine.org/en/stable/classes/class_tween.html) — Interruptible `create_tween()` motion for cell-to-cell feedback while logical `Vector2i` state updates immediately.
- [Using TileMaps](https://docs.godotengine.org/en/stable/tutorials/2d/using_tilemaps.html) — TileMapLayer workflows for painting walls/targets while keeping puzzle truth in a separate grid dictionary.
- [Saving games](https://docs.godotengine.org/en/stable/tutorials/io/saving_games.html) — Persist level progress, stars, and mid-puzzle snapshots without relying on scene reload as save.
- [Using InputEvent](https://docs.godotengine.org/en/stable/tutorials/inputs/inputevent.html) — Device-agnostic click/drag/action routing for forgiving grid selection and move commits.
- [Resources](https://docs.godotengine.org/en/stable/tutorials/scripting/resources.html) — Store layouts and starting piece sets as `.tres`/`Resource` data editable in the Inspector.
- [Physics introduction](https://docs.godotengine.org/en/stable/tutorials/physics/physics_introduction.html) — RigidBody sleep, layers, and integration hooks for physics-driven puzzle pieces.
- [Idle and Physics Processing](https://docs.godotengine.org/en/stable/tutorials/scripting/idle_and_physics_processing.html) — Why win/soft-lock checks belong on move commits, not every `_process` frame.
- [Data preferences](https://docs.godotengine.org/en/stable/tutorials/best_practices/data_preferences.html) — Prefer integer grid keys (`Vector2i`) and explicit dictionaries over SceneTree-as-truth.
- [Runtime file loading and saving](https://docs.godotengine.org/en/stable/tutorials/io/runtime_file_loading_and_saving.html) — `FileAccess`/`user://` patterns for custom level JSON and editor export packs.
- [JSON](https://docs.godotengine.org/en/stable/classes/class_json.html) — Serialize compact puzzle boards when splitting `Vector2` fields for portable level files.

### Related Skills

#### Prerequisites
- [godot-input-handling](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-input-handling/SKILL.md) — Buffered InputEvent/action maps so grid clicks and directional moves stay device-agnostic and forgiving.
- [godot-signal-architecture](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-signal-architecture/SKILL.md) — Safe `state_changed` / `level_complete` wiring so UI and VFX stay decoupled from grid truth.
- [godot-project-foundations](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-project-foundations/SKILL.md) — Autoload/project layout baselines before stacking undo managers, savers, and level packs.

#### Complements
- [godot-tweening](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-tweening/SKILL.md) — Deeper Tween composition when cell moves, match clears, and resets need interruptible juice without logic races.
- [godot-save-load-systems](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-save-load-systems/SKILL.md) — Progress ownership, versioning, and threaded loads beyond per-level JSON snapshots.
- [godot-tilemap-mastery](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-tilemap-mastery/SKILL.md) — TileMapLayer painting, custom data, and terrain patterns that visualize walls while scripts own solvability.
- [godot-navigation-pathfinding](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-navigation-pathfinding/SKILL.md) — Broader Navigation/A* stacks when puzzle hints outgrow a single `AStarGrid2D` region.
- [godot-2d-physics](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-2d-physics/SKILL.md) — RigidBody sleep, layers, and queries for physics puzzles that still need deterministic settle/win checks.
- [godot-state-machine-advanced](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-state-machine-advanced/SKILL.md) — Phase FSMs (observe → move → resolve → win) when puzzles mix animation locks with input gates.
- [godot-camera-systems](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-camera-systems/SKILL.md) — Camera2D/3D framing and unproject helpers for perspective/world-space puzzle overlays.
- [godot-ui-containers](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-ui-containers/SKILL.md) — Minimal undo/reset HUD layouts that stay non-intrusive during non-verbal tutorials.

#### Downstream / consumers
- [godot-monte-carlo-balancer](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-monte-carlo-balancer/SKILL.md) — Sample solvability, move-count distributions, and soft-lock rates so level packs stay fair as mechanics combine.
- [godot-procedural-generation](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-procedural-generation/SKILL.md) — Generate candidate boards that still pass this skill's validators, undo constraints, and win-condition contracts.
- [godot-genre-roguelike](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-genre-roguelike/SKILL.md) — Consumes grid/undo patterns when dungeon runs embed discrete puzzle rooms or locked-door logic.

#### Master
- [godot-master](https://github.com/thedivergentai/gd-agentic-skills/blob/main/skills/godot-master/SKILL.md) — Library router and mirrored module entry for discovering this genre skill beside sibling domains.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Godot Genre Puzzle AI skill do?

Expert blueprint for puzzle games including undo systems (Command pattern for state reversal), grid-based logic (Sokoban-style mechanics), non-verbal tutorials (teach through level design), win condition checking, state management, and visual feedback (instant confirmation of valid moves). Use for logic puzzles, physics puzzles, or match-3 games. Trigger keywords: puzzle_game, undo_system, command_pattern, grid_logic, non_verbal_tutorial, state_management.

Why use Godot Genre Puzzle on TypingMind?

Because you install it once and use it with any model. Godot Genre Puzzle is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Godot Genre Puzzle in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/thedivergentai/GD-Agentic-Skills/tree/main/skills/godot-genre-puzzle. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Godot Genre Puzzle?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Godot Genre Puzzle?

As many as you like. As long as a model supports skills, you can use Godot Genre Puzzle with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Godot Genre Puzzle AI skill free?

Yes. It is published on GitHub by thedivergentai under the LGPL-3.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇