Fix articulation and schema documentation CI failures

[kellyguo11]

Description

The test_external_force_on_multiple_bodies_at_position test asserted only the world-Z component of root angular velocity. The applied offset wrench can produce rotation on another axis, especially when body-frame wrenches are transformed, so the articulation can rotate while the Z component remains below the threshold.

Check the total angular-speed magnitude instead, preserving the existing 0.1 threshold. Apply the same assertion to the OVPhysX, PhysX, and Newton test mirrors.

This addresses the recurring CUDA failure observed in the test_articulation job, where the world-Z angular velocity was 0.004370182752609253.

Related to #7458.

No runtime dependencies are added.

Type of change

• Bug fix (non-breaking change which fixes an issue)

Release backport

• Backport this pull request to the active release branch after it merges into develop

Screenshots

Not applicable; this is a test-only change.

Validation

• uv run --no-project python -m compileall -q on all three modified test modules
• uv run --no-project python tools/changelog/cli.py check codex-changelog-base --include-worktree
• uvx pre-commit run --all-files
• The simulator-backed CUDA case requires Linux GPU CI and will be requested with run-ci

Checklist

• I have read and understood the contribution guidelines
• I have run the pre-commit checks with ./isaaclab.sh --format
• I have made corresponding changes to the documentation (not applicable: test-only change)
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works (updated the existing mirrored regression tests)
• I have added a changelog fragment under source/<pkg>/changelog.d/ for every touched package (do not edit CHANGELOG.rst or bump extension.toml — CI handles that)
• I have added my name to the CONTRIBUTORS.md or my name already exists there

Documentation CI follow-up

The docs workflow also exposed 14 duplicate object registrations for schema fragment classes documented on both isaaclab.sim and isaaclab.sim.schemas. Keep the top-level public API rendering but mark its alias copies :no-index:, leaving the canonical schema page as the only Sphinx cross-reference target.

Failure: Build Latest Docs.

A full local Sphinx build completed without the original duplicate-object warnings. Remaining local warnings were import failures for project packages unavailable in the macOS validation environment; the pushed Linux docs workflow provides the authoritative warning-free result.

[kellyguo11]

run-ci

[isaaclab-review-bot]

Isaac Lab Review Bot

The PR consistently updates the mirrored Newton, PhysX, and OVPhysX articulation tests to check total root angular speed rather than only the world-Z component, while preserving the existing threshold and adding appropriate .skip changelog fragments.

• Design and architecture: The axis-agnostic assertion is applied identically across all three backend test mirrors. No production architecture or shared helper is changed; the main tradeoff is reduced axis specificity, which matches the stated intent because the response axis depends on link-frame transformations.
• API: No public API or extension-facing behavior changes. The patch only reads the existing root_ang_vel_w test data, and each touched source package includes the required non-user-visible changelog fragment.
• Implementation: Using torch.linalg.vector_norm on the three-component angular velocity directly tests whether the articulation rotates regardless of response axis. Keeping the 0.1 threshold preserves the prior minimum-response criterion, while broadening the accepted response from Z-axis rotation to rotation on any axis.

No blocking issues. No inline issue met the actionable-evidence threshold; the assessment above records the review feedback.

Automated review; human maintainers own approval decisions.

[greptile-apps]

Greptile Summary

Updates mirrored articulation tests to measure total root angular speed instead of only the world-Z component, preventing valid rotation on another axis from failing the test.

• Applies the same vector-norm assertion to the Newton, OVPhysX, and PhysX test suites.
• Adds package-scoped skip changelog fragments for all three touched backends.

Confidence Score: 5/5

The test-only change appears safe to merge.

The updated assertions continue to require substantial articulation rotation while avoiding an invalid dependency on a particular world-space axis, and the mirrored backend tests remain consistent.

Important Files Changed

Filename	Overview
source/isaaclab_newton/test/assets/test_articulation.py	Replaces the world-Z angular-velocity assertion with total angular-speed magnitude in the Newton wrench-response test.
source/isaaclab_ov/test/assets/test_articulation.py	Applies the same axis-independent rotation assertion to the OVPhysX test mirror.
source/isaaclab_physx/test/assets/test_articulation.py	Applies the same axis-independent rotation assertion to the PhysX test mirror.

_{Reviews (1): Last reviewed commit: "Fix articulation wrench response asserti..." | Re-trigger Greptile}

[kellyguo11]

run-ci