MCP server providing tools for Unreal Engine editor automation, asset management, actor manipulation, animation authoring, and AI behavior graph editing
Critical failures across naming, descriptions, and schemas. 9 of 20 tools (45%) have empty input schemas with no parameters defined, automatic 0 for schema. Tools 9-20 lack actionable descriptions (under 20 chars or generic placeholders). Naming is inconsistent: 'spawn_actor' vs 'delete_actor' vs 'asset.get_mesh_bounds' (dot notation breaks verb_noun convention). No output schemas documented. Error handling absent. Only tools 1-8 have partial schemas; tools 9-20 are stubs marked 'private' with no implementation visible. This server is not production-ready.
Add a behavior definition to a smart object (private tool)
Add a slot to a smart object (private tool)
Rename multiple assets in a single operation
Get the bounding box and sphere bounds of a mesh asset
Get collision information for a mesh asset
Auto-arrange behavior tree nodes (private tool)
Clone a behavior tree subtree (private tool)
9 of 20 tools (45%) have empty input schemas with no parameters defined. Tools 9-20 marked 'private' show Input: {} with no fields.
Descriptions for tools 9-20 are either missing or trivial (under 20 chars: 'private tool').
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 34 | 2026-07-28+ | v2 |
Configure a smart object slot (private tool)
Delete an actor from the editor world (requires confirmation)
Add a smart object slot (gameplay tool)
Add behavior to a smart object slot (gameplay tool)
Remove a smart object slot (gameplay tool)
Set smart object slot properties (gameplay tool)
Generate a behavior tree diagram (private tool)
Modify a property on an asset
Remove a behavior definition from a smart object (private tool)
Remove a slot from a smart object (private tool)
Set the transform (location, rotation, scale) of an actor
Set the visibility of an actor
Spawn an actor of a given class at a specified location and rotation in the editor world
Naming inconsistency: tools use both verb_noun (spawn_actor, delete_actor) and dot notation (asset.get_mesh_bounds, gameplay.add_smart_object_slot). Dot notation breaks LLM parsing of intent. Recommend flat namespace with prefixes: get_mesh_bounds, add_smart_object_slot.
No output schemas documented for any tool. LLMs cannot plan downstream calls or extract required fields. Tools 1-8 have input schemas but no return type documentation.
delete_actor requires 'confirmed' boolean but lacks error guidance if confirmation fails. No recovery path documented. Per pattern:recovery-guide, errors must tell LLM what to do next.