What problem does it solve?
Building Procedural Content Generation graphs in Unreal Engine by hand is slow and error-prone, and the PCG Python API has non-obvious pin labels, protected properties, and save/reload requirements that cause silent failures. This Skill provides verified patterns for creating, wiring, and generating PCG graphs without trial-and-error.
Core Features & Use Cases
- Graph Authoring: Create PCG Graph assets, add nodes (Surface Sampler, Static Mesh Spawner, filters), and connect pins with correct labels via
add_edge.
- Volume Generation: Place a PCGVolume, assign a graph with
set_graph, run generate(force=True), and verify output through ISM instance counts.
- Pitfall Prevention: Built-in guidance for known traps such as the Volume Sampler
"Volume" pin, Vector-typed voxel_size, protected graph property, and mandatory save-then-reload cycles.
- Use Case: Ask the agent to scatter static meshes across a landscape; it creates the graph, wires Surface Sampler to Static Mesh Spawner, configures mesh entries, places a PCGVolume, and generates instances.
Quick Start
Use the pcg skill to create a PCG graph at /Game/PCG that scatters cube meshes with a Surface Sampler and Static Mesh Spawner, then place a PCGVolume and generate.