> For the complete documentation index, see [llms.txt](https://amazed-moose-studio.gitbook.io/mindcraftai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://amazed-moose-studio.gitbook.io/mindcraftai/readme.md).
# MindCraft AI
**MindCraft AI is a powerful GOAP (Goal-Oriented Action Planning) system for Unreal Engine 5.5, built in C++ for maximum efficiency.** Designed to seamlessly integrate with **MetaHuman technology**, it enables your human-like NPCs to **think, plan, and act naturally**, making them behave with lifelike intelligence and adaptability. Whether it's **decision-making, tactical planning, or dynamic responses to the environment**, MindCraft AI brings **unparalleled realism** to your NPCs.
GOAP is not just for games—similar **goal-driven AI systems** power real-world **autonomous space exploration vehicles** like NASA’s **Curiosity rover and lunar landers**. These spacecraft **analyze objectives, plan sequences of actions, and adapt to unexpected challenges**—just like MindCraft AI enables NPCs to do. By leveraging GOAP’s **dynamic action selection and replanning capabilities**, your NPCs will respond intelligently to changing conditions, prioritize objectives, and operate with the same kind of autonomy seen in **cutting-edge robotics and space technology**.
Goal-Oriented Action Planning is planning architecture designed for real-time control of autonomous character behavior in games. Originally implemented GOAP for [F.E.A.R](https://web.archive.org/web/20230912145018/http://en.wikipedia.org/wiki/F.E.A.R.). This A.I. architecture simultaneously powered Monolith's [Condemned: Criminal Origins](https://web.archive.org/web/20230912145018/http://www.gamerankings.com/htmlpages2/926309.asp?q=condemned).
A **Finite State Machine (FSM)** dictates exactly how an AI should behave in every situation, following predefined rules and transitions. In contrast, a **planning system like GOAP (Goal-Oriented Action Planning)** provides the AI with **goals and available actions**, allowing it to dynamically determine the best sequence of actions to achieve those goals.
FSMs are **procedural**, requiring explicit state transitions for every possible scenario, whereas **GOAP is declarative**, enabling AI to **evaluate, plan, and adapt** in real-time. This flexibility allows for the emergence of **complex, human-like behaviors** in NPCs, making them more intelligent, unpredictable, and lifelike.
For deep explanation of GOAP system itself please refer to [**this**](https://web.archive.org/web/20230912173042/https://alumni.media.mit.edu/~jorkin/aiide05OrkinJ.pdf) **and** [**this**](https://web.archive.org/web/20230912145018/https://alumni.media.mit.edu/~jorkin/goap.html)**.**
And of course please see this [short video](https://youtu.be/4c0yApZS1eM) of NPC crew unloading cargo with the help of MIndCraft AI.
## R.E.A.C.T.O.R System
**R.E.A.C.T.O.R. is the ultimate enhancement for GOAP-driven AI, bringing NPCs to life with dynamic personality-driven decision-making.** Unlike traditional AI systems that rely solely on logic and pre-scripted behaviors, R.E.A.C.T.O.R. introduces **Reflexes, Endurance, Accuracy, Courage, Temperament, Observation, and Resolve**, creating NPCs that **think, adapt, and react realistically**. Each attribute dynamically influences **GOAP action planning**, allowing NPCs to make choices based on their physical and psychological state—**a cautious scout may avoid combat, while a fearless soldier charges forward**. Integrated seamlessly with GOAP, R.E.A.C.T.O.R. ensures that **NPCs are not just goal-driven but also personality-driven, leading to organic, emergent behaviors** where no two AI characters act exactly the same. This makes AI encounters feel fresh, challenging, and deeply immersive, perfect for next-generation gaming experiences.
You might find it useful to join our [Discord ](https://discord.gg/auydUABXzG)for any additional questions.
### Jump right in
