Use when writing, reviewing, or refactoring SwiftUI code for iOS or macOS, including state management and `@Observable` data flow, view composition and invalidation/performance, lists and `ForEach` identity, environment usage, localization, animations, Liquid Glass adoption, migrating soft-deprecated APIs, or Instruments `.trace` capture/analysis for hangs, hitches, CPU hotspots, or excessive view updates.
All-time #737Trending #1305Hot #6350First seen Jan 24, 2026
Expert SwiftUI code review, refactoring, and implementation guidance with iOS 26+ and macOS support.
Covers state management, view composition, performance optimization, and modern SwiftUI patterns across iOS 15+ through iOS 26+ Flags deprecated APIs and suggests modern replacements; gates version-specific features with #available and fallbacks Addresses lists, layouts, animations, accessibility, Swift Charts, and platform-specific patterns (macOS scenes, Liquid Glass) Enforces correctness rules for property wrappers, ForEach identity, and view diffing; presents performance improvements as optional suggestions
Similar popular skills
Related neighbors and high-traction skills in the same topics — useful to compare before installing.
The swiftui-expert-skill is a specialized tool for SwiftUI development and performance profiling. It provides a robust set of references and Python scripts for recording and analyzing Xcode Instruments trace files. Security analysis confirms that the skill is safe to use, employing standard Python libraries and local system utilities (xctrace) with secure command execution patterns. No malicious code, obfuscation, or unauthorized data access behaviors were detected.
Pick a template based on target kind — the SwiftUI template populates the SwiftUI lane on any real device: a physical iOS/iPadOS device or the host Mac. The only exception is the iOS Simulator, where the SwiftUI lane comes back empty — switch to --template "Time Profiler" in that case (still gives Time Profiler + Hangs + Animation Hitches). Always check --list-devices: simulators kind → Time Profiler; devices kind (real devices and the host Mac) → default SwiftUI. Full decision table in references/trace-recording.md.
Start the recording. For agent-driven sessions where the user says "I'll tell you when I'm done", start in the background and use a stop-file:
``bash python3 "${SKILLDIR}/scripts/recordtrace.py" \ --device "<name|udid>" --attach "<AppName>" \ --stop-file /tmp/stop-trace --output ~/Desktop/session.trace `` For interactive sessions, just tell the user to press Ctrl+C when done.
Signal stop — when the user says they've finished exercising the app, touch /tmp/stop-trace. The script cleanly SIGINTs xctrace and waits up to 60s for finalisation.
Analyse the resulting trace (flow into the "Trace-driven improvement" workflow below).
Trigger whenever the user's request references a .trace file. A target SwiftUI source file is optional — if given, cite specific lines; if not, recommend where to look based on view names and symbols the trace already reveals.
Full reference: references/trace-analysis.md. Summary of the composition pattern:
Scope the analysis. Ask yourself: does the user want the whole trace, or a slice?
- "focus on X / after X / between X and Y / during X" → resolve to a window first (see step 2). - No scoping cue → analyse the whole trace.
Resolve a window (only if the user scoped). The parser exposes two discovery modes:
``bash # Find a log that marks the start/end of the region of interest: python3 "${SKILLDIR}/scripts/analyzetrace.py" --trace <path> \ --list-logs --log-message-contains "loaded feed" --log-limit 5 # Or list ossignpost intervals (paired begin/end), filterable by name: python3 "${SKILLDIR}/scripts/analyzetrace.py" --trace <path> \ --list-signposts --signpost-name-contains "ImageDecode" ` Both modes accept --window STARTMS:ENDMS to scope discovery. Pick the timems (for logs) or startms/endms (for signposts) that match the user's description. Build a window like --window 10400:11700`.
Interpret with references/trace-analysis.md — key diagnostics:
- mainrunningcoveragepct inside each correlation (<25% = blocked; ≥75% = CPU-bound). - swiftui-causes.topsources reveals why updates keep happening — high-edge-count sources like UserDefaultObserver.send() or wide EnvironmentWriter entries are structural invalidation bugs. Fixing one often collapses many downstream hot views.
When a specific view shows as expensive, ask who's invalidating it. Use --fanin-for "<view name>" to get the ranked list of source nodes driving the updates.
Optionally ground in source. If the user pointed at a file, read it and match view names / user-code symbols against identifiers there. If not, recommend which files to open based on the view names SwiftUI reported.
Return a prioritised plan. Cite evidence (coverage %, hot symbol, overlapping view, log timestamp, cause-graph edges) and route each recommendation to a Topic Router reference.
Only edit code if the user asked for edits.
Topic Router
Consult the reference file for each topic relevant to the current task:
Topic
Reference
State management
references/state-management.md
View composition
references/view-structure.md
Performance
references/performance-patterns.md
Lists and ForEach
references/list-patterns.md
Layout
references/layout-best-practices.md
Sheets and navigation
references/sheet-navigation-patterns.md
ScrollView, scroll position, and scroll geometry
references/scroll-patterns.md
Focus management
references/focus-patterns.md
Animations (basics)
references/animation-basics.md
Animations (transitions)
references/animation-transitions.md
Animations (advanced)
references/animation-advanced.md
Accessibility
references/accessibility-patterns.md
Swift Charts
references/charts.md
Charts accessibility
references/charts-accessibility.md
Image optimization
references/image-optimization.md
Liquid Glass (iOS 26+)
references/liquid-glass.md
macOS scenes
references/macos-scenes.md
macOS window styling
references/macos-window-styling.md
macOS views
references/macos-views.md
Text patterns
references/text-patterns.md
Localization
references/localization.md
Deprecated API lookup
references/latest-apis.md
Handling soft-deprecated APIs
references/soft-deprecation.md
Previews
references/previews.md
Instruments trace analysis
references/trace-analysis.md
Instruments trace recording
references/trace-recording.md
Correctness Checklist
These are hard rules -- violations are always bugs:
@State properties are private
@Binding only where a child modifies parent state
Passed values never declared as @State or @StateObject (they ignore updates)
@StateObject for view-owned objects; @ObservedObject for injected
iOS 17+: @State with @Observable; @Bindable for injected observables needing bindings
ForEach uses stable identity (never .indices/\.offset; id outlives the view and isn't derived from mutable content)
Constant number of views per ForEach element; List rows are unary
No closures stored in custom @Environment/@FocusedValue keys
Custom @Entry default values are stable (no Model()/Date()/UUID() expressions)
.animation(_:value:) always includes the value parameter
@FocusState properties are private
No redundant @FocusState writes inside tap gesture handlers on .focusable() views
iOS 26+ APIs gated with #available and fallback provided
import Charts present in files using chart types
Previews use self-contained mock data; no dependency on live services or network
References
references/latest-apis.md -- Read first for every task. Deprecated-to-modern API transitions (iOS 15+ through iOS 26+)
references/state-management.md -- Property wrappers, data flow, @Observable migration
references/soft-deprecation.md -- How to behave with soft-deprecated APIs (when to migrate, scoping rule, don't migrate during unrelated edits)
references/trace-analysis.md -- Parse Instruments .trace files via scripts/analyze_trace.py; interpret main-thread coverage, high-severity SwiftUI updates, hitch narratives, and map findings back to source files
references/trace-recording.md -- Record a new trace via scripts/record_trace.py: attach to a running app, launch one fresh, or capture a manually-stopped session; supports stop-file for agent-driven flows