decentraland / builder / 30509469048

52%

master: 52%

Build Type

Pull #3454

github

Committed by RocioCM

Commit Message

feat: pose wearables for auto-generated thumbnails by category

Upper body and skin wearables were screenshot in their authored bind
pose (T-pose), which hides the wearable's details. They are now captured
in a static pose and from the front camera, using the same offscreen
renderer mechanism that already gives skin-like categories their
front-facing camera angle.

- src/poses/pose_nicer_idle.glb: mesh-less single-keyframe animation on
  the standard Avatar_* armature (new top-level poses/ folder, aliased
  like icons/ and images/)
- getScreenshot: optional 'pose' option that writes the pose keyframes
  directly onto the model's matching transform nodes before capturing
- getItemData: categories with a pose render their import-time thumbnail
  with the offscreen engine instead of the wearable-preview iframe
- THUMBNAIL_POSE_BY_CATEGORY maps upper_body and skin to the pose;
  changing the category in the details step regenerates the thumbnail
  when the camera angle or pose differs (as skin already did)

Adding a pose for another category = one GLB + one map entry.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Pull Request Pull Request #3454: feat: pose wearables for auto-generated thumbnails by category

Coverage Stats

2303 of 5583 branches covered (41.25%)

Branch coverage included in aggregate %.

6 of 32 new or added lines in 3 files covered. (18.75%)

1 existing line in 1 file now uncovered.

7282 of 12746 relevant lines covered (57.13%)

34.4 hits per line