Object Call Override Create

Part of Avatar Graph, Avatech AI's pack for turning a single character portrait into a real-time interactive 2D avatar - blinking, lip sync, driven by Blender's Python engine (bpy) plus Avatech's own "Shape Flow" runtime for live shape-key control. Avatar Graph's README only spotlights a handful of nodes you'd actually build a workflow around - Create Mesh Layer, Create Shape Flow, Avatar Main Output. Object Call Override Create didn't make that list. It's one of a large batch of ObjectCall_* nodes exposing Blender's bpy.types.Object API roughly one method per node, which is a code-generated surface, not a curated one - no individual description anywhere in the repo.

The Blender feature underneath: library overrides. When an object is linked in from an external .blend file, you normally can't edit it directly - your options are leave it untouched or make_local() it entirely, cutting the tie to the source file for good. Library overrides split the difference: object.override_create(remap_local_usages=False) creates a shallow, editable override of the linked object, so you can tweak specific properties locally while the object stays connected to (and can still receive updates from) its source library. remap_local_usages (default false) controls whether existing local references to the original linked object get redirected to point at the new override instead.

Whether this ever comes up in a real Avatar Graph workflow depends on whether your pipeline links in shared rig/mesh assets from a library file rather than building everything from the pack's own mesh nodes - if you're doing the latter (the common case, going by the README's own templates), you probably never touch this node.

Schema-wise: inputs are BPY_OBJ (optional per the schema, but the node needs one wired in to do anything) and remap_local_usages, a plain BOOLEAN - a genuinely normal, no-fuss ComfyUI toggle, unlike a lot of its siblings in this node family that fall back to placeholder types for enum/object arguments. Output is BPY_OBJ, the same object handle passed straight through for chaining into more Object-Call nodes.

Installing it means installing the whole pack. ComfyUI Manager: search "Avatar Graph." Manually: cd ComfyUI/custom_nodes && git clone https://github.com/avatechai/avatar-graph-comfyui, then cd avatar-graph-comfyui && pip install -r requirements.txt, restart. The requirement worth planning for: bpy needs Python 3.10.x exactly - following the README's advice to build a dedicated conda create --name comfyui python=3.10 env rather than bolting this onto your existing Flux/SDXL setup will save you a fight. Restart with --enable-cors-header (--force-fp16 too on macOS) so the live avatar preview via Avatech's editor works.

Troubleshooting: bpy refusing to import at startup because ComfyUI's Python isn't exactly 3.10.x is the most likely wall you'll hit, and it's a common, recurring category of pain for Blender-Python-based ComfyUI nodes generally. Forgetting --enable-cors-header doesn't break execution - you can still export a .glb/.gltf off Avatar Main Output - you just lose the live preview panel, which is deliberate per the pack's FAQ. This node has no documentation of its own in the repo, so if library overrides are a new concept to you, Blender's own manual pages on override workflows and Object.override_create() are where the real explanation lives - this node is just a thin pass-through to that API.