State Trees are Unreal's data-driven alternative to Behavior Trees, built for both single-actor AI and Mass Entity crowds at scale. The architecture is clever: tasks and conditions are lightweight structs instead of UObjects, and execution context gets reconstructed each tick from persistent instance data, making parallel evaluation safe. You get hierarchical states with evaluators that run globally before transitions, plus a schema system that validates what data your tree can touch at edit time. The instance data pattern takes getting used to (all your mutable state lives in a separate struct), but it pays off when you're running thousands of entities. If you're migrating from Behavior Trees or building AI that needs to scale beyond a few dozen actors, this is the modern path forward.
npx skills add https://github.com/quodsoler/unreal-engine-skills --skill ue-state-trees