orkas-ai/orkas/b1f384166705
- community
Checking that the SKILL.md file is still in the repository.
Install
npx vskill@latest install orkas-ai/orkas/b1f384166705SKILL.md
SwiftUI Dev
Use this skill for SwiftUI development, architecture, structure, performance, and Apple native app profiling. It combines:
- SwiftUI view refactor and Observation/MV guidance.
- SwiftUI code-first performance audit.
- Native macOS/iOS Time Profiler CLI workflow with
xctrace,atos, and included scripts.
Do not use this skill for web performance, server profiling, generic PR review, product planning, or non-Apple UI frameworks. For ordinary implementation work without SwiftUI architecture/performance/native profiling concerns, use the development skill.
Route The Work
| User intent | Read | |
|---|---|---|
Refactor SwiftUI View structure, split large body, review Observation boundaries | SwiftUI refactor | |
| Decide whether MV, framework-native state, or a ViewModel is justified | MV patterns | |
| SwiftUI page is slow, scrolling janks, or body updates too often | SwiftUI performance | |
| Build a dependency/update mental model before deeper profiling | WWDC23 performance model | |
| Inspect current SwiftUI Instrument lanes or Cause & Effect evidence | SwiftUI Instruments workflow and SwiftUI timeline guide | |
| Diagnose a main-thread hang or run-loop stall | App hangs guide | |
| Record/analyze macOS or iOS native Time Profiler traces from CLI, symbolicate and rank hotspots through the standard Orkas Skill Runner | Native trace workflow |
If the user provides only symptoms, start with code-first SwiftUI review. Ask for trace/screenshots only when code review is inconclusive or the user explicitly wants trace analysis.
Required Inputs
For SwiftUI code review:
- Target view or feature code.
- Data flow:
@State,@Binding,@Environment,@Observable,@Query, services, models. - Symptoms and reproduction steps.
- Device/OS/build configuration when performance is involved.
For native profiling:
- App name and binary path, or existing
.tracebundle. - Attach PID or launch binary path.
- Slow interaction to exercise during capture.
- Capture duration, defaulting to 90 seconds.
- Matching symbols / binary and permission to launch or attach.
Proceed with explicit assumptions for code-only reviews. Stop before profiling if required tools or permissions are missing.
Core Workflow
- Classify the task: SwiftUI structure/refactor, SwiftUI performance, or native trace profiling.
- Start code-first for SwiftUI: inspect structure, data flow, Observation ownership, identity, layout, and body work before requesting heavier tooling.
- Use Instruments evidence when available: treat traces/screenshots/CSV as evidence, not as the whole answer.
- Make targeted recommendations: order issues by likely impact and confidence.
- Verify with the same scenario: compare before/after CPU, frame drops, memory peak, sample counts, or the specific user-visible symptom when data exists.
Boundaries
- Do not invent metrics, trace findings, sample counts, call stacks, device settings, or benchmark results.
- Do not run
xctraceunless the user has provided a local target, trace, or explicit permission and the environment is macOS with Xcode tools. - Do not interpret stale symbols as real hotspots. Binary, trace, and load address must match.
- Do not introduce ViewModels by default. Prefer SwiftUI-native state/data flow unless existing code or user requirements justify a view model.
- Do not change business logic during refactor unless the user explicitly asks.
- If source files must be edited, follow the existing project style and verify with relevant build/tests/profiling.
Output Format
Use this concise structure unless a reference template is more specific:
For native trace analysis, use the report structure in the native trace workflow.
Quality Checklist
Before finalizing, verify:
- The task is Apple-native or SwiftUI-specific.
- SwiftUI structure recommendations preserve clear data ownership and framework-native state flow.
- Findings cite code, trace evidence, screenshots, or clearly labeled assumptions.
- Recommendations are targeted, not generic "optimize everything" advice.
- Refactors preserve behavior unless requested otherwise.
- Profiling commands, trace path, binary path, load address, and confidence are reported when CLI profiling is used.
- The user gets a concrete verification plan.
Security
- Security scanPassed, 96 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingresources/builtin/marketplace/skills/b1f384166705/SKILL.md
Popularity
- Repository stars
- 2,173
- Repository forks
- 90
- 7-day trend score
- 0
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/orkas-ai/orkas/b1f384166705)