State Machine logo

State Machine

Community
jame581
state-machine

Use when implementing state machines in Godot — enum-based, node-based, and resource-based FSM patterns with trade-offs

Overview

Publisherjame581
RepositoryGodotPrompter
Skill namestate-machine
Stars
808
Forks
47
Bundled files
1
LicenseMIT
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.

  • 1 bundled files

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

  • Open source

    Published by jame581 on GitHub. Read the source before you install it.

Installation

Install the State Machine 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.
    https://github.com/jame581/GodotPrompter/tree/master/skills/state-machine
  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/jame581/GodotPrompter.git /tmp/GodotPrompter
mkdir -p .claude/skills
cp -r /tmp/GodotPrompter/skills/state-machine .claude/skills/state-machine
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable State Machine 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 State Machine 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 State Machine 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.

State Machines in Godot 4.3+

Choose the right FSM pattern for your complexity level. All examples target Godot 4.3+ with no deprecated APIs.

Related skills: player-controller for movement state integration, ai-navigation for AI state patterns, resource-pattern for resource-based state configuration, animation-system for AnimationTree states driven by FSM, dialogue-system for dialogue flow as a state machine, ability-system for caster state gating (casting/stunned), limboai for the LimboAI addon's HSM (BTState) if you need a behavior tree alongside your FSM, beehave for a GDScript-only BT alternative.

When to reach for an addon: This skill covers the built-in FSM patterns (enum, node-based, resource-based). If your agent needs a full behavior tree, see limboai (C++ + HSM, Godot 4.6+) or beehave (pure GDScript, Godot 4.1+) instead.


1. Approach Comparison

ApproachComplexityBest For
Enum-BasedLowSimple objects, fewer than 5 states
Node-BasedMediumCharacters with complex behavior
Resource-BasedHighData-driven or editor-configurable AI

2. Approach 1: Enum-Based (Simplest)

Use when you have a small number of states and no significant enter/exit logic.

GDScript

gdscript
extends CharacterBody2D

enum State { IDLE, PATROL, CHASE, ATTACK }

@export var patrol_range: float = 200.0
@export var chase_range: float = 300.0
@export var attack_range: float = 50.0
@export var speed: float = 80.0

var current_state: State = State.IDLE
var patrol_target: Vector2 = Vector2.ZERO

@onready var player: Node2D = get_tree().get_first_node_in_group("player")


func _physics_process(delta: float) -> void:
	match current_state:
		State.IDLE:
			_state_idle()
		State.PATROL:
			_state_patrol()
		State.CHASE:
			_state_chase()
		State.ATTACK:
			_state_attack()

	move_and_slide()


func _state_idle() -> void:
	velocity = Vector2.ZERO
	if _player_in_range(chase_range):
		current_state = State.CHASE
	elif randf() < 0.005:
		patrol_target = global_position + Vector2(randf_range(-patrol_range, patrol_range), 0.0)
		current_state = State.PATROL


func _state_patrol() -> void:
	var direction := (patrol_target - global_position)
	if direction.length() < 4.0:
		current_state = State.IDLE
		return
	velocity = direction.normalized() * speed
	if _player_in_range(chase_range):
		current_state = State.CHASE


func _state_chase() -> void:
	if not is_instance_valid(player):
		current_state = State.IDLE
		return
	if _player_in_range(attack_range):
		current_state = State.ATTACK
		return
	if not _player_in_range(chase_range):
		current_state = State.PATROL
		return
	velocity = (player.global_position - global_position).normalized() * speed


func _state_attack() -> void:
	velocity = Vector2.ZERO
	if not _player_in_range(attack_range):
		current_state = State.CHASE


func _player_in_range(range: float) -> bool:
	if not is_instance_valid(player):
		return false
	return global_position.distance_to(player.global_position) <= range

C# Equivalent

csharp
using Godot;

public partial class SimpleEnemy : CharacterBody2D
{
    private enum State { Idle, Patrol, Chase, Attack }

    [Export] public float PatrolRange { get; set; } = 200f;
    [Export] public float ChaseRange  { get; set; } = 300f;
    [Export] public float AttackRange { get; set; } = 50f;
    [Export] public float Speed       { get; set; } = 80f;

    private State _currentState = State.Idle;
    private Vector2 _patrolTarget = Vector2.Zero;
    private Node2D _player;

    public override void _Ready()
    {
        _player = GetTree().GetFirstNodeInGroup("player") as Node2D;
    }

    public override void _PhysicsProcess(double delta)
    {
        switch (_currentState)
        {
            case State.Idle:   StateIdle();   break;
            case State.Patrol: StatePatrol(); break;
            case State.Chase:  StateChase();  break;
            case State.Attack: StateAttack(); break;
        }
        MoveAndSlide();
    }

    private void StateIdle()
    {
        Velocity = Vector2.Zero;
        if (PlayerInRange(ChaseRange))
        {
            _currentState = State.Chase;
        }
        else if (GD.Randf() < 0.005f)
        {
            _patrolTarget = GlobalPosition + new Vector2(GD.RandRange(-PatrolRange, PatrolRange), 0f);
            _currentState = State.Patrol;
        }
    }

    private void StatePatrol()
    {
        var direction = _patrolTarget - GlobalPosition;
        if (direction.Length() < 4f) { _currentState = State.Idle; return; }
        Velocity = direction.Normalized() * Speed;
        if (PlayerInRange(ChaseRange)) _currentState = State.Chase;
    }

    private void StateChase()
    {
        if (!IsInstanceValid(_player)) { _currentState = State.Idle; return; }
        if (PlayerInRange(AttackRange)) { _currentState = State.Attack; return; }
        if (!PlayerInRange(ChaseRange)) { _currentState = State.Patrol; return; }
        Velocity = (_player.GlobalPosition - GlobalPosition).Normalized() * Speed;
    }

    private void StateAttack()
    {
        Velocity = Vector2.Zero;
        if (!PlayerInRange(AttackRange)) _currentState = State.Chase;
    }

    private bool PlayerInRange(float range) =>
        IsInstanceValid(_player) && GlobalPosition.DistanceTo(_player.GlobalPosition) <= range;
}

When to upgrade away from enum-based:

  • Enter/exit logic starts duplicating across state methods
  • Animation sync requires explicit enter/exit hooks
  • The match/switch block grows beyond ~100 lines

3. Approach 2: Node-Based (Recommended for Characters)

Each state is its own node. The StateMachine node delegates input and process calls to whichever state is active, and states trigger transitions by name.

Scene Tree

Player (CharacterBody2D)
└── StateMachine (Node)
    ├── Idle  (State)
    ├── Run   (State)
    ├── Jump  (State)
    └── Attack (State)

State Base Class

GDScript (state.gd)

gdscript
class_name State
extends Node

## Populated by StateMachine._ready()
var entity: CharacterBody2D
var state_machine: StateMachine


## Called when this state becomes active.
func enter() -> void:
	pass


## Called when this state is deactivated.
func exit() -> void:
	pass


## Mirrors _process. Return a state name string to transition, or "" to stay.
func update(delta: float) -> String:
	return ""


## Mirrors _physics_process. Return a state name string to transition, or "".
func physics_update(delta: float) -> String:
	return ""


## Mirrors _unhandled_input.
func handle_input(event: InputEvent) -> String:
	return ""

C# (State.cs)

csharp
using Godot;

public partial class State : Node
{
    /// Populated by StateMachine._Ready()
    public CharacterBody2D Entity { get; set; }
    public StateMachine StateMachine { get; set; }

    public virtual void Enter() { }
    public virtual void Exit() { }
    public virtual string Update(double delta) => string.Empty;
    public virtual string PhysicsUpdate(double delta) => string.Empty;
    public virtual string HandleInput(InputEvent @event) => string.Empty;
}

StateMachine Class

GDScript (state_machine.gd)

gdscript
class_name StateMachine
extends Node

@export var initial_state: State

var current_state: State
var states: Dictionary = {}


func _ready() -> void:
	for child in get_children():
		if child is State:
			states[child.name] = child
			child.entity = owner as CharacterBody2D
			child.state_machine = self

	if initial_state:
		current_state = initial_state
		current_state.enter()


func _unhandled_input(event: InputEvent) -> void:
	var next := current_state.handle_input(event)
	if next:
		transition_to(next)


func _process(delta: float) -> void:
	var next := current_state.update(delta)
	if next:
		transition_to(next)


func _physics_process(delta: float) -> void:
	var next := current_state.physics_update(delta)
	if next:
		transition_to(next)


func transition_to(state_name: String) -> void:
	if not states.has(state_name):
		push_error("StateMachine: unknown state '%s'" % state_name)
		return
	current_state.exit()
	current_state = states[state_name]
	current_state.enter()

C# (StateMachine.cs)

csharp
using System.Collections.Generic;
using Godot;

public partial class StateMachine : Node
{
    [Export] public State InitialState { get; set; }

    public State CurrentState { get; private set; }
    private readonly Dictionary<string, State> _states = new();

    public override void _Ready()
    {
        foreach (var child in GetChildren())
        {
            if (child is State state)
            {
                _states[state.Name] = state;
                state.Entity = Owner as CharacterBody2D;
                state.StateMachine = this;
            }
        }

        if (InitialState != null)
        {
            CurrentState = InitialState;
            CurrentState.Enter();
        }
    }

    public override void _UnhandledInput(InputEvent @event)
    {
        var next = CurrentState.HandleInput(@event);
        if (!string.IsNullOrEmpty(next)) TransitionTo(next);
    }

    public override void _Process(double delta)
    {
        var next = CurrentState.Update(delta);
        if (!string.IsNullOrEmpty(next)) TransitionTo(next);
    }

    public override void _PhysicsProcess(double delta)
    {
        var next = CurrentState.PhysicsUpdate(delta);
        if (!string.IsNullOrEmpty(next)) TransitionTo(next);
    }

    public void TransitionTo(string stateName)
    {
        if (!_states.TryGetValue(stateName, out var next))
        {
            GD.PushError($"StateMachine: unknown state '{stateName}'");
            return;
        }
        CurrentState.Exit();
        CurrentState = next;
        CurrentState.Enter();
    }
}

Concrete Example: IdleState

GDScript (idle_state.gd)

gdscript
class_name IdleState
extends State


func enter() -> void:
	entity.get_node("AnimationPlayer").play("idle")


func physics_update(delta: float) -> String:
	if not entity.is_on_floor():
		return "Jump"
	if Input.get_axis("move_left", "move_right") != 0.0:
		return "Run"
	return ""


func handle_input(event: InputEvent) -> String:
	if event.is_action_pressed("jump") and entity.is_on_floor():
		return "Jump"
	if event.is_action_pressed("attack"):
		return "Attack"
	return ""

4. Approach 3: Resource-Based (Data-Driven)

Use when designers need to configure states in the Godot Inspector without modifying code.

StateData Resource

gdscript
class_name StateData
extends Resource

@export var state_name: String = ""
@export var animation_name: String = ""
@export var move_speed: float = 0.0
@export var can_transition_to: Array[String] = []

Export an Array[StateData] on your AI controller. Designers populate each entry in the Inspector — no code changes needed to tune behavior or add states. The runtime reads can_transition_to to validate transitions and picks animation_name / move_speed for each active state.

csharp
using Godot;

[GlobalClass]
public partial class StateData : Resource
{
    [Export] public string StateName { get; set; } = string.Empty;
    [Export] public string AnimationName { get; set; } = string.Empty;
    [Export] public float MoveSpeed { get; set; } = 0f;
    [Export] public Godot.Collections.Array<string> CanTransitionTo { get; set; } = new();
}

Attach an Array[StateData] export on your AI controller class ([Export] public Godot.Collections.Array<StateData> States). At runtime, look up the active StateData by StateName and read AnimationName / MoveSpeed to drive behavior; use CanTransitionTo to guard TransitionTo calls.


5. Hierarchical and Parallel State Machines

When a flat FSM grows beyond ~8 states or spans multiple concerns (movement + combat + animation), split into hierarchical machines (states own sub-state machines, e.g. OnGround containing Idle/Walk/Run) or parallel machines (independent FSMs for movement, combat, animation running side-by-side). Both keep state counts additive instead of multiplicative.

See references/hierarchical-and-parallel.md for full scene trees, HierarchicalState base class, parallel-machine character example, and a "which to choose" comparison table — GDScript and C# for each.


6. Decision Flowchart

Start
  │
  ▼
Fewer than 5 states?
  ├─ Yes ──────────────────────────────────► Enum-Based
  └─ No
       │
       ▼
     Multiple independent concerns
     (movement + combat + animation)?
       ├─ Yes ──────────────────────────────► Parallel State Machines
       └─ No
            │
            ▼
          States naturally nest
          (sub-states within states)?
            ├─ Yes ────────────────────────► Hierarchical State Machine
            └─ No
                 │
                 ▼
               Designers need to configure
               states in the Inspector?
                 ├─ Yes ──────────────────► Resource-Based
                 └─ No ──────────────────► Node-Based

7. Implementation Checklist

  • Chose the approach that matches actual complexity (enum / node / resource)
  • Every state has explicit enter() and exit() methods (or equivalent)
  • All transitions are named explicitly — no implicit fallthrough between states
  • Animations are started in enter() and cleaned up in exit() where needed
  • No circular transition loops that could cause infinite recursion in a single frame
  • Flat FSM is replaced with hierarchical or parallel when states exceed ~8 or span multiple concerns
  • Parallel state machines don't modify the same state (e.g., both setting velocity) — one concern per machine

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 State Machine AI skill do?

Use when implementing state machines in Godot — enum-based, node-based, and resource-based FSM patterns with trade-offs

Why use State Machine on TypingMind?

Because you install it once and use it with any model. State Machine 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 State Machine in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/jame581/GodotPrompter/tree/master/skills/state-machine. 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 State Machine?

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 State Machine?

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

Is the State Machine AI skill free?

Yes. It is published on GitHub by jame581 under the MIT 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 👇