macos-spatial-metal-engineer
Creates Metal shaders, builds instanced 3D rendering pipelines, optimizes GPU performance, and implements spatial computing interactions for macOS and visionOS/Vision Pro. Handles Compositor Services stereo frame streaming, RemoteImmersiveSpace setup, gaze and pinch gesture recognition, and GPU-based physics for graph layout. Use when the user asks about Metal shaders, .metal files, 3D rendering on macOS, GPU programming, visionOS or Vision Pro development, spatial computing, RealityKit, SceneKit, AR, augmented reality, instanced rendering, render pipeline optimization, or stereoscopic output.
90
88%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Passed
No known issues
Quality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is an excellent skill description that thoroughly covers specific capabilities across Metal shader development, GPU rendering pipelines, and visionOS spatial computing. It uses proper third-person voice, lists concrete actions, and includes a comprehensive 'Use when...' clause with abundant natural trigger terms spanning both technical and user-friendly vocabulary. The description is distinctive enough to avoid conflicts while being comprehensive enough to catch all relevant user queries.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: creating Metal shaders, building instanced 3D rendering pipelines, optimizing GPU performance, implementing spatial computing interactions, handling Compositor Services stereo frame streaming, RemoteImmersiveSpace setup, gaze and pinch gesture recognition, and GPU-based physics for graph layout. | 3 / 3 |
Completeness | Clearly answers both 'what does this do' (creates Metal shaders, builds pipelines, optimizes GPU performance, implements spatial computing interactions, etc.) AND 'when should Claude use it' with an explicit 'Use when...' clause listing numerous trigger scenarios. | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural terms users would say: Metal shaders, .metal files, 3D rendering, GPU programming, visionOS, Vision Pro, spatial computing, RealityKit, SceneKit, AR, augmented reality, instanced rendering, render pipeline optimization, stereoscopic output. These cover both technical and colloquial variations. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive niche combining Metal/GPU programming with visionOS/Vision Pro spatial computing. The specific technologies (Compositor Services, RemoteImmersiveSpace, .metal files) make it very unlikely to conflict with other skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a strong, highly actionable skill with executable code examples, concrete performance targets, and a well-structured workflow with explicit validation checkpoints. Its main weakness is that it's a large monolithic document that could benefit from splitting detailed code examples into separate reference files, and some sections could be slightly more concise. The Compositor Services pseudocode disclaimer is a thoughtful touch for an evolving API.
Suggestions
Split the large code blocks (Metal rendering pipeline, Compositor Services, spatial interaction, compute shader) into separate reference files and link to them from the main SKILL.md to improve progressive disclosure.
Trim the Core Mission section and inline comments that restate what the code already makes obvious to improve token efficiency.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is fairly long with substantial code examples that are useful but could be tightened. Some sections like the Core Mission bullet points and the Advanced Capabilities paragraph add moderate value, but the overall length (~200+ lines) includes details Claude could infer. The Compositor Services pseudocode with caveats about SDK evolution is reasonable but slightly verbose. | 2 / 3 |
Actionability | The skill provides fully executable Swift and Metal shader code, concrete struct definitions, specific API calls, and precise performance targets (90 fps, 25k nodes, ≤11.1 ms frame time, ≤1 GB memory). The Metal compute kernel is complete and copy-paste ready, and the rendering pipeline code is concrete with real Metal API usage. | 3 / 3 |
Workflow Clarity | The workflow section has four clearly sequenced steps, each with explicit validation checkpoints including specific tools (Metal System Trace, Instruments, Metal GPU Debugger) and concrete pass/fail criteria (sustained 90 fps, frame time ≤11.1 ms, memory ≤1 GB). The progression from pipeline setup → rendering → Vision Pro integration → optimization is logical with feedback loops implied at each checkpoint. | 3 / 3 |
Progressive Disclosure | The content is well-structured with clear headers and sections, but it's essentially a monolithic document. The large code blocks for the rendering pipeline, compositor, interaction system, and compute shader could be split into separate reference files. The Advanced Capabilities section is appropriately condensed as a summary, but no external file references are provided for deeper dives. | 2 / 3 |
Total | 10 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 10 / 11 Passed | |
- Repository
- OpenRoster-ai/awesome-agents
- Commit
010799b
- 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.