Zum Inhalt springen
Medium 1 für Eintrag Character Rig AI Toolset

Beschreibung

Inspect and edit IK retargeting, skeletal structures, sockets, skin weights and Control Rig pose previews through native Unreal MCP tools.

What it does

• Inspect and edit retargeter operation stacks, rig chains and native controllers.

• Read skeletal hierarchies, reference transforms and sockets.

• Inspect dominant bone weights and perform rigid binding and mesh composition.

• Measure live PIE actor, bone, socket and morph state.

• Build pose stages, calibrate controls, check poses and capture review images.

Requirements

Unreal Engine 5.8, Windows 64-bit editor. Enable ControlRig, ToolsetRegistry, PythonScriptPlugin, IKRig, MeshModelingToolset, GeometryScripting, EditorScriptingUtilities, SequencerScripting, LevelSequenceEditor, SkeletalMeshModelingTools, ModelContextProtocol. Connect an HTTP MCP-compatible client to the native editor server. Editor only; source and documentation included. No project or third-party asset packs are included.

Documentation and support

https://github.com/ITemre/CharacterRigAIToolset

https://github.com/ITemre/CharacterRigAIToolset/issues

Limitations

• Meshes, animations and Control Rig assets are supplied by the buyer. No character, animation pack or third-party content is included.

• Walk and jog generation are experimental Manny-specific workflows, not a general animation generator. They require /Game/Characters/Mannequins/Meshes/SKM_Manny_Simple.

• Pose stages use /Game/PoseLab/L_PoseLab and save their level and sequence. Use a disposable project when first learning the workflow.

• Bone rebuilding can remove morph data stored only in render data. Work on duplicates.

• Pose collision and balance checks are estimates and require visual review. A successful check does not establish animation quality.

Validation

Built and checked in UE 5.8.4 on Windows. 194 distinct internal tests covered with real Manny/Control Rig assets, retargeting fixtures and PIE. The 14 capture tests passed again after the final preview update. Fixed Manny walk/jog experiments were generated and verified. Expected rejection diagnostics are included where applicable. See the public validation notes for fixture scope and limitations.

Enthaltene Formate