GameReadyGoap 2.1.0

<img src="https://github.com/Joy-less/GameReadyGoap/blob/main/Assets/Icon.jpg?raw=true" width=256/>

Game Ready Goap

An easy-to-use implementation of GOAP (Goal-Oriented Action Planning) to control game characters in C#.

Features

• Simple and performant, made for game development
• Expressive with minimal boilerplate
• Get as close as possible to "best-effort" goals

Usage

First, create an agent with initial states, goals and actions:

GoapAgent Agent = new() {
    // These describe the current state of your agent (character).
    States = new() {
        ...
    },
    // These are the states your agent is trying to achieve.
    Goals = [
        ...
    ],
    // These are the ways your agent can change their states.
    Actions = [
        ...
    ],
};

Then, finding a plan is easy:

Agent.FindPlan(); // or Agent.FindPlan(Goal);

Executing plans is also easy:

Plan.Execute(Agent, Action => {
    ...
});

Example

A farmer is balancing tending to his crops with resting. He can farm to increase his crop health, which requires energy, or sleep to increase his energy.

GoapAgent Farmer = new() {
    States = new() {
        ["Energy"] = 100,
        ["CropHealth"] = 0,
    },
    Goals = [
        new GoapGoal("TendToCrops") {
            Objectives = [
                new GoapCondition() {
                    State = "CropHealth",
                    Comparison = GoapComparison.GreaterThanOrEqualTo,
                    Value = 100,
                    BestEffort = true,
                },
            ],
        },
    ],
    Actions = [
        new GoapAction("Farm") {
            Effects = [
                new GoapEffect() {
                    State = "CropHealth",
                    Operation = GoapOperation.IncreaseBy,
                    Value = 20,
                },
                new GoapEffect() {
                    State = "Energy",
                    Operation = GoapOperation.DecreaseBy,
                    Value = 30,
                },
            ],
            Requirements = [
                new GoapCondition() {
                    State = "Energy",
                    Comparison = GoapComparison.GreaterThanOrEqualTo,
                    Value = 30,
                },
            ],
        },
        new GoapAction("Sleep") {
            Effects = [
                new GoapEffect() {
                    State = "Energy",
                    Operation = GoapOperation.IncreaseBy,
                    Value = 5,
                },
            ],
        },
    ],
};
Farmer.FindPlan();

We get 15 actions which bring us to our goal:

Tips

Avoid near-impossible goals

Instead of aiming to set the player's health = 0, aim for ⇐ 0. Best-effort plans are slow to find because plans that reach the goal are always prioritised.

Avoid lots of actions

The more actions an agent has, the longer it will take to find a plan.

Know the limitations

GOAP requires a lot of trial and error to use successfully. If configured poorly, it may result in unpredictable plans that are feasible but not believable.

Special Thanks

• F.E.A.R. for creating the GOAP algorithm.
• This Is Vini for explaining the GOAP algorithm.
• SimpleGOAP for guidance when implementing the action planner.