CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

scenekit

Maintains existing SceneKit scenes, nodes, materials, cameras, lights, animation, physics, particles, assets, shaders, and SwiftUI hosting. Use for SceneKit maintenance; route new 3D apps, major redesigns, USD/USDZ pipelines, and migration planning to RealityKit.

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

78%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured maintenance skill: an lean overview with real, clearly scoped one-level-deep reference bundles, executable DON'T/DO Swift examples, and an unusually specific review checklist. The recurring gaps are duplicated reference pointers across sections, descriptive rather than executable workflow steps, and the absence of an explicit validate-fix-retry feedback loop.

Suggestions

Consolidate the reference pointers: merge "Route by Task" and "References" (or keep one section) so each of references/core-implementation.md and references/scenekit-patterns.md is described once with a single authoritative scope list, removing the 2-3 duplicated links.

Add an explicit feedback loop to the Workflow, e.g. after step 5: "If rendering, memory, or frame-time checks fail, re-inspect the node hierarchy and physics ownership from step 2 before proceeding" — this would lift workflow_clarity to the 5 anchor.

Make Workflow steps 2-4 executable by pointing each step at the concrete recipe in the references (e.g. step 2 → coordinate-conversion and hierarchy inspection snippets in references/core-implementation.md) instead of descriptive phrasing like "explicit lifecycle ownership".

DimensionReasoningScore

Conciseness

The body assumes Claude's competence with no conceptual padding (the WWDC 2025 deprecation note is load-bearing routing guidance, not gratuitous version trivia), but references/core-implementation.md and references/scenekit-patterns.md are each described and linked 2-3 times across "Route by Task" and "References" with overlapping scope lists — trimmable duplication that keeps it below "every token earns its place".

4 / 5

Actionability

Common Mistakes contains complete, executable Swift (SCNCamera setup with pointOfView, SCNPhysicsShape with a simplified SCNBox, applyForce(_, asImpulse:)) and the checklist items are concretely verifiable ("Lights limited to 8 per node; attenuationEndDistance set on point/spot lights"), but Workflow steps 2-4 ("Inspect scene ownership, node hierarchy...", "with explicit lifecycle ownership") describe rather than instruct, leaving minor gaps relative to fully copy-paste-ready coverage.

4 / 5

Workflow Clarity

The 5-step Workflow has a clear sequence with an explicit verification step ("Verify model loading, device rendering, memory, frame time, and SwiftUI/UIKit teardown") plus a ~20-item Review Checklist, but there is no error-recovery feedback loop (no "if verification fails, fix and re-check" guidance), which the 5 anchor requires; this is not a destructive/batch skill, so no cap applies.

4 / 5

Progressive Disclosure

The entrypoint is a lean ~110-line overview with a TOC, and both referenced files (references/core-implementation.md, references/scenekit-patterns.md) exist as real one-level-deep bundles with clearly signaled scope descriptions ("custom geometry, shader modifiers, constraints, morph targets, serialization, hit testing, LOD, and renderer delegates"); the ~1088 lines of implementation detail are appropriately split out, matching the clear-overview anchor.

5 / 5

Total

17

/

20

Passed

Description

87%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: it states a comprehensive what, an explicit Use-for trigger clause, and boundary routing to RealityKit that sharply reduces conflict risk. The only weaknesses are the umbrella verb "Maintains" in place of several distinct concrete actions, and missing synonyms/file extensions (.scn, .dae, SCNView) that would improve trigger matching.

DimensionReasoningScore

Specificity

"Maintains existing SceneKit scenes, nodes, materials, cameras, lights, animation, physics, particles, assets, shaders, and SwiftUI hosting" covers the domain comprehensively with concrete objects, but the actions reduce to a single umbrella verb ("Maintains") plus "route", not the multiple distinct concrete actions of the 5 anchor; well above the 1-2-action coverage of the 3 anchor.

4 / 5

Completeness

Explicitly answers both: what ("Maintains existing SceneKit scenes, nodes, materials, ... and SwiftUI hosting") and when ("Use for SceneKit maintenance"), and additionally specifies when not to use it ("route new 3D apps, major redesigns, USD/USDZ pipelines, and migration planning to RealityKit") — matching the anchor-5 pattern of concrete, explicit trigger guidance rather than the merely adequate 'when' of the 4 anchor.

5 / 5

Trigger Term Quality

Natural terms users would say are present ("SceneKit", "3D apps", "RealityKit", "USD/USDZ", "SwiftUI", "nodes", "physics", "shaders"), giving good keyword coverage, but common variations like file extensions (.scn, .dae) and API class names (SCNView, SCNScene) are missing, so it falls short of the 5 anchor's synonym and extension coverage.

4 / 5

Distinctiveness Conflict Risk

"Use for SceneKit maintenance" is a clear niche trigger, and the description actively disambiguates against its closest conflicting skill ("route new 3D apps, major redesigns, USD/USDZ pipelines, and migration planning to RealityKit"), giving minimal conflict risk.

5 / 5

Total

18

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

thienqo/ios-skills

Path

skills/scenekit/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.