fix: use deepClone for Skin._skinMatrices to prevent undefined after clone

[cptbtptpbcptdtptp]

Summary

• Change @ignoreClone to @deepClone for Skin._skinMatrices property
• This ensures cloned Skin objects have their own Float32Array for skin matrices instead of undefined

Problem

When cloning a Skin object, _skinMatrices was decorated with @ignoreClone. According to the clone logic in CloneManager.cloneProperty(), when CloneMode.Ignore is set, the property is skipped entirely during cloning. This results in the cloned Skin object having _skinMatrices as undefined, which causes rendering issues when the cloned SkinnedMeshRenderer tries to update skin matrices.

Solution

Use @deepClone instead to create a deep copy of the Float32Array, ensuring each cloned Skin instance has its own properly initialized skin matrices array.

Test plan

• Verify Skin cloning creates independent _skinMatrices
• Verify cloned SkinnedMeshRenderer renders correctly
• Run existing unit tests

🤖 Generated with Claude Code

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 91d3e096-537b-4647-8b21-634c30b15d50

📥 Commits

Reviewing files that changed from the base of the PR and between 0597e61 and 76f6a98.

📒 Files selected for processing (1)

• packages/core/src/mesh/Skin.ts

---

Walkthrough

The _skinMatrices internal field in the Skin class has its clone behavior annotation updated from @ignoreClone to @deepClone, changing how the Float32Array buffer is handled during object duplication.

Changes

Cohort / File(s)	Summary
Clone Annotation Update packages/core/src/mesh/Skin.ts	Modified the _skinMatrices field's clone annotation from @ignoreClone to @deepClone, ensuring the buffer is deep copied during cloning operations rather than ignored.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

Poem

🐰 A copy, a clone, a buffer so true,
From ignore to deep, we make something new,
The skinned matrices dance with their kin,
No longer forgotten—they'll follow along in! ✨

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately summarizes the main change: switching the clone annotation for Skin._skinMatrices from @ignoreClone to @deepClone to prevent shared references between cloned instances.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 77.12%. Comparing base (0597e61) to head (76f6a98).
⚠️ Report is 6 commits behind head on dev/2.0.

Additional details and impacted files

@@             Coverage Diff             @@
##           dev/2.0    #2941      +/-   ##
===========================================
+ Coverage    77.07%   77.12%   +0.04%     
===========================================
  Files          899      899              
  Lines        98229    98229              
  Branches      9683     9693      +10     
===========================================
+ Hits         75706    75755      +49     
+ Misses       22355    22306      -49     
  Partials       168      168              

Flag	Coverage Δ
unittests	77.12% <100.00%> (+0.04%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[augmentcode]

🤖 Augment PR Summary

Summary: Fixes Skin cloning to avoid shared skinning matrix buffers between the original and its clone.

Changes: Switches Skin._skinMatrices from @ignoreClone to @deepClone so clones get their own Float32Array copy.

_{🤖 Was this summary useful? React with 👍 or 👎}

[augmentcode]

Review completed. 1 suggestion posted.

Comment augment review to trigger a new review at any time.

[augmentcode]

Consider adding a small regression test that Skin.clone() produces an independent _skinMatrices (i.e., not the same Float32Array reference), since this change is specifically fixing shared-state cloning bugs.

Severity: low

_{🤖 Was this useful? React with 👍 or 👎, or 🚀 if it prevented an incident/outage.}

[GuoLei1990]

总结

将 Skin._skinMatrices 的 clone 装饰器从 @ignoreClone 改为 @deepClone，修复 clone 后 _skinMatrices 为 undefined 导致渲染异常的问题。改动一行，方向正确。

问题

• [P2] packages/core/src/mesh/Skin.ts — @deepClone 对 Float32Array 会创建独立副本，这在语义上是正确的（每个 Skin 实例需要独立的 skinMatrices buffer）。但需要确认 _skinMatrices 的初始化时机：如果它是在 update 时按需分配的（lazy init），那么 clone 时源对象的 _skinMatrices 可能尚未初始化（仍为初始默认值），@deepClone 会正确复制该初始值。反之，如果源对象已有一个正确 size 的 Float32Array，clone 后 target 拿到的是一份快照，后续如果 target 的 bone 数量不同（通过 retarget），这个 buffer size 可能不匹配。建议确认 clone 后是否有 resize 逻辑覆盖此场景。

• [P3] — PR 描述中的 test plan 全是 unchecked（- [ ]），建议补齐或标注已通过。

简化建议

无，改动已经是最小的。