swiftui-performance

Diagnose and fix SwiftUI rendering bottlenecks through code review, Instruments profiling, and targeted remediation. Covers view invalidation storms, unstable list identity, expensive body computations, layout thrash, and image decoding issues with concrete code examples and fixes Includes step-by-step Instruments profiling workflow using the SwiftUI template to identify high body-evaluation counts and CPU hotspots Explains identity and lifetime mechanics, lazy loading patterns, and @Observable granular tracking to narrow observation scope and reduce unnecessary re-renders Provides a review checklist and common mistakes reference (formatters in body, GeometryReader in lazy containers, blocking MainActor, etc.)

INSTALLATION
npx skills add https://github.com/dpearson2699/swift-ios-skills --skill swiftui-performance
Run in your project or agent environment. Adjust flags if your CLI version differs.

SKILL.md

SwiftUI Performance

Audit SwiftUI view performance end-to-end, from instrumentation and baselining to root-cause analysis and concrete remediation steps.

Contents

  • [Workflow Decision Tree](#workflow-decision-tree)
  • [1. Code-First Review](#1-code-first-review)
  • [2. Guide the User to Profile](#2-guide-the-user-to-profile)
  • [3. Analyze and Diagnose](#3-analyze-and-diagnose)
  • [4. Remediate](#4-remediate)
  • [Common Code Smells (and Fixes)](#common-code-smells-and-fixes)
  • [5. Verify](#5-verify)
  • [Outputs](#outputs)
  • [Instruments Profiling](#instruments-profiling)
  • [Identity and Lifetime](#identity-and-lifetime)
  • [Lazy Loading Patterns](#lazy-loading-patterns)
  • [State and Observation Optimization](#state-and-observation-optimization)
  • [Common Mistakes](#common-mistakes)
  • [Review Checklist](#review-checklist)
  • [References](#references)

Workflow Decision Tree

  • If the user provides code, start with "Code-First Review."
  • If the user only describes symptoms, ask for minimal code/context, then do "Code-First Review."
  • If code review is inconclusive, go to "Guide the User to Profile" and ask for a trace or screenshots.

1. Code-First Review

Collect:

  • Target view/feature code.
  • Data flow: state, environment, observable models.
  • Symptoms and reproduction steps.

Focus on:

  • View invalidation storms from broad state changes.
  • Unstable identity in lists (id churn, UUID() per render).
  • Top-level conditional view swapping (if/else returning different root branches).
  • Heavy work in body (formatting, sorting, image decoding).
  • Layout thrash (deep stacks, GeometryReader, preference chains).
  • Large images without downsampling or resizing.
  • Over-animated hierarchies (implicit animations on large trees).

Provide:

  • Likely root causes with code references.
  • Suggested fixes and refactors.
  • If needed, a minimal repro or instrumentation suggestion.

2. Guide the User to Profile

Explain how to collect data with Instruments:

  • Use the SwiftUI template in Instruments.
  • Profile a Release build on a real device when possible.
  • Reproduce the exact interaction (scroll, navigation, animation).
  • Capture SwiftUI timeline and Time Profiler.
  • Export or screenshot the relevant lanes and the call tree.

Ask for:

  • Trace export or screenshots of SwiftUI lanes + Time Profiler call tree.
  • Device/OS/build configuration.

3. Analyze and Diagnose

Prioritize likely SwiftUI culprits:

  • View invalidation storms from broad state changes.
  • Unstable identity in lists (id churn, UUID() per render).
  • Top-level conditional view swapping (if/else returning different root branches).
  • Heavy work in body (formatting, sorting, image decoding).
  • Layout thrash (deep stacks, GeometryReader, preference chains).
  • Large images without downsampling or resizing.
  • Over-animated hierarchies (implicit animations on large trees).

Summarize findings with evidence from traces/logs.

4. Remediate

Apply targeted fixes:

  • Narrow state scope (@State/@Observable closer to leaf views).
  • Stabilize identities for ForEach and lists.
  • Move heavy work out of body (precompute, cache, @State).
  • Use equatable() or value wrappers for expensive subtrees.
  • Downsample images before rendering.
  • Reduce layout complexity or use fixed sizing where possible.

Common Code Smells (and Fixes)

Look for these patterns during code review.

Expensive formatters in body

var body: some View {

    let number = NumberFormatter() // slow allocation

    let measure = MeasurementFormatter() // slow allocation

    Text(measure.string(from: .init(value: meters, unit: .meters)))

}

Prefer cached formatters in a model or a dedicated helper:

final class DistanceFormatter {

    static let shared = DistanceFormatter()

    let number = NumberFormatter()

    let measure = MeasurementFormatter()

}

Computed properties that do heavy work

var filtered: [Item] {

    items.filter { $0.isEnabled } // runs on every body eval

}

Prefer precompute or cache on change:

@State private var filtered: [Item] = []

// update filtered when inputs change

Sorting/filtering in body or ForEach

// DON'T: sorts or filters on every body evaluation

ForEach(items.sorted(by: sortRule)) { item in Row(item) }

ForEach(items.filter { $0.isEnabled }) { item in Row(item) }

Prefer precomputed, cached collections with stable identity. Update on input change, not in body.

Unstable identity

ForEach(items, id: \.self) { item in

    Row(item)

}

Avoid id: \.self for non-stable values; use a stable ID.

Top-level conditional view swapping

var content: some View {

    if isEditing {

        editingView

    } else {

        readOnlyView

    }

}

Prefer one stable base view and localize conditions to sections/modifiers (for example inside toolbar, row content, overlay, or disabled). This reduces root identity churn and helps SwiftUI diffing stay efficient.

Image decoding on the main thread

Image(uiImage: UIImage(data: data)!)

Prefer decode/downsample off the main thread and store the result.

Broad dependencies in observable models

@Observable class Model {

    var items: [Item] = []

}

var body: some View {

    Row(isFavorite: model.items.contains(item))

}

Prefer granular view models or per-item state to reduce update fan-out.

5. Verify

Ask the user to re-run the same capture and compare with baseline metrics.

Summarize the delta (CPU, frame drops, memory peak) if provided.

Outputs

Provide:

  • A short metrics table (before/after if available).
  • Top issues (ordered by impact).
  • Proposed fixes with estimated effort.

Instruments Profiling

Use the SwiftUI instrument template in Xcode (Cmd+I to profile). Key instruments: SwiftUI View Body (body evaluation counts), SwiftUI View Properties (state change tracking), Time Profiler, and Hangs.

Add Self._printChanges() in debug builds to log which property triggered a view update:

var body: some View {

    #if DEBUG

    let _ = Self._printChanges()  // "MyView: @self, _count changed."

    #endif

    Text("Count: \(count)")

}

See references/optimizing-swiftui-performance-instruments.md for the full profiling workflow.

Identity and Lifetime

Structural Identity vs Explicit Identity

SwiftUI assigns every view an identity used to track its lifetime, state, and animations.

  • Structural identity (default): determined by the view's position in the view hierarchy. SwiftUI uses the call-site location in body to distinguish views.
  • Explicit identity: you assign with .id(_:) modifier or ForEach(items, id: \.stableID).
// Structural identity: SwiftUI knows these are different views by position

VStack {

    Text("First")   // position 0

    Text("Second")  // position 1

}

How Identity Tracks View Lifetime

When a view's identity changes, SwiftUI treats it as a new view:

  • All @State is reset.
  • onAppear fires again.
  • Animations may restart.
  • Transition animations play (if defined).

When identity stays the same, SwiftUI updates the existing view in place, preserving state and providing smooth transitions.

AnyView and Identity Reset

AnyView erases type information, forcing SwiftUI to fall back to less efficient diffing:

// DON'T: AnyView destroys type identity

func makeView(for item: Item) -> AnyView {

    if item.isPremium {

        return AnyView(PremiumRow(item: item))

    } else {

        return AnyView(StandardRow(item: item))

    }

}

// DO: use @ViewBuilder to preserve structural identity

@ViewBuilder

func makeView(for item: Item) -> some View {

    if item.isPremium {

        PremiumRow(item: item)

    } else {

        StandardRow(item: item)

    }

}

AnyView also prevents SwiftUI from detecting which branch changed, causing full subtree replacement instead of targeted updates.

Ternary Modifiers Preserve Structural Identity

if/else in a view builder creates _ConditionalContent — two separate view branches with distinct identities. When the condition changes, SwiftUI destroys one branch and creates the other, resetting all @State.

For toggling modifiers on the same view, use a ternary expression instead:

// DON'T: if/else creates two separate Text views with different identities

if isHighlighted {

    Text(title).foregroundStyle(.yellow)

} else {

    Text(title).foregroundStyle(.primary)

}

// DO: ternary keeps one Text view, just changes the modifier value

Text(title)

    .foregroundStyle(isHighlighted ? .yellow : .primary)

This preserves the view's identity (and its state) across the condition change, and SwiftUI can animate the transition smoothly.

Use if/else when the view type itself differs between branches. Use ternary when only a property or modifier changes.

id() Modifier Impacts

The .id() modifier assigns explicit identity. Changing the value destroys and recreates the view:

// DON'T: UUID() changes every render, destroying and recreating the view each time

ScrollView {

    LazyVStack {

        ForEach(items) { item in

            Row(item: item)

                .id(UUID())  // kills performance -- new identity every render

        }

    }

}

// DO: use a stable identifier

ForEach(items) { item in

    Row(item: item)

        .id(item.stableID)  // identity only changes when the item actually changes

}

Intentional .id() change is useful for resetting state (e.g., .id(selectedTab) to reset a scroll position when switching tabs).

Lazy Loading Patterns

LazyVStack and LazyHStack

Lazy stacks only create views for items currently visible on screen. Off-screen items are not evaluated until scrolled into view.

ScrollView {

    LazyVStack {

        ForEach(items) { item in

            ItemRow(item: item)

        }

    }

}

Key behaviors:

  • Views are created lazily but not destroyed when scrolled off screen (they remain in memory).
  • onAppear fires when the view first enters the visible area.
  • onDisappear fires when it leaves, but the view is still alive.

LazyVGrid and LazyHGrid

Use lazy grids for multi-column layouts:

// Adaptive: as many columns as fit with minimum width

let columns = [GridItem(.adaptive(minimum: 150))]

ScrollView {

    LazyVGrid(columns: columns) {

        ForEach(photos) { photo in

            PhotoThumbnail(photo: photo)

        }

    }

}

// Fixed: exact number of equal columns

let fixedColumns = [

    GridItem(.flexible()),

    GridItem(.flexible()),

    GridItem(.flexible()),

]

When to Use Lazy vs Eager Stacks

Scenario

Use

< 50 items

VStack / HStack (eager is fine)

50-100 items

Either works; prefer Lazy if items are complex

100 items

LazyVStack / LazyHStack (required for performance)

Always-visible content

VStack (no benefit to lazy)

Scrollable lists

LazyVStack inside ScrollView, or List

Important: Do not nest GeometryReader inside lazy containers. It forces eager measurement and defeats lazy loading. Use .onGeometryChange (iOS 16+) instead.

State and Observation Optimization

@Observable Granular Tracking

@Observable (Observation framework, iOS 17+) tracks property access at the per-property level. A view only re-evaluates when properties it actually read in body change:

@Observable class UserProfile {

    var name: String = ""

    var avatarURL: URL?

    var biography: String = ""

}

// This view ONLY re-renders when `name` changes -- not when

// biography or avatarURL change, because it only reads `name`

struct NameLabel: View {

    let profile: UserProfile

    var body: some View {

        Text(profile.name)

    }

}

This is a significant improvement over ObservableObject + @Published, which invalidates all observing views when any published property changes.

Avoiding Observation Scope Pollution

If a view reads many properties from an @Observable model in body, it re-renders when any of those properties change. Push reads into child views to narrow the scope:

// DON'T: reads name, email, avatar, and settings in one body

struct ProfileView: View {

    let model: ProfileModel

    var body: some View {

        VStack {

            Text(model.name)           // tracks name

            Text(model.email)          // tracks email

            AsyncImage(url: model.avatar) // tracks avatar

            SettingsForm(model.settings)  // tracks settings

        }

    }

}

// DO: split into child views so each only tracks what it reads

struct ProfileView: View {

    let model: ProfileModel

    var body: some View {

        VStack {

            NameRow(model: model)      // only tracks name

            EmailRow(model: model)     // only tracks email

            AvatarView(model: model)   // only tracks avatar

            SettingsForm(model: model) // only tracks settings

        }

    }

}

Computed Properties for Derived State

Use computed properties on @Observable models to derive state without introducing extra stored properties that widen observation scope:

@Observable class ShoppingCart {

    var items: [CartItem] = []

    // Views reading `total` only re-render when `items` changes

    var total: Decimal {

        items.reduce(0) { $0 + $1.price * Decimal($1.quantity) }

    }

}

Common Mistakes

  • Profiling Debug builds. Debug builds include extra runtime checks and disable optimizations, producing misleading perf data. Profile Release builds on a real device.
  • Observing an entire model when only one property is needed. Break large @Observable models into focused ones, or use computed properties/closures to narrow observation scope.
  • **Using GeometryReader inside ScrollView items.** GeometryReader forces eager sizing and defeats lazy loading. Prefer .onGeometryChange (iOS 16+) or measure outside the lazy container.
  • **Calling DateFormatter() or NumberFormatter() inside body.** These are expensive to create. Make them static or move them outside the view.
  • Animating non-equatable state. If SwiftUI cannot determine equality, it redraws every frame. Conform state to Equatable, then use .animation(_:value:) for simple value-bound changes or .animation(_:body:) for narrower modifier-scoped implicit animation.
  • **Large flat List without identifiers.** Use id: or make items Identifiable so SwiftUI can diff efficiently instead of rebuilding the entire list.
  • **Unnecessary @State wrapper objects.** Wrapping a simple value type in a class for @State defeats value semantics. Use plain @State with structs.
  • **Blocking MainActor with synchronous I/O.** File reads, JSON parsing of large payloads, and image decoding should happen off the main actor. Use Task.detached or a custom actor.

Review Checklist

  • No DateFormatter/NumberFormatter allocations inside body
  • Large lists use Identifiable items or explicit id:
  • @Observable models expose only the properties views actually read
  • Heavy computation is off MainActor (image processing, parsing)
  • GeometryReader is not inside a LazyVStack/LazyHStack/List
  • Implicit animations use .animation(_:value:) for value-bound changes or .animation(_:body:) for narrower modifier scope
  • No synchronous network/file I/O on the main thread
  • Profiling done on Release build, real device
  • @Observable view models are @MainActor-isolated; types crossing concurrency boundaries are Sendable

References

BrowserAct

Let your agent run on any real-world website

Bypass CAPTCHA & anti-bot for free. Start local, scale to cloud.

Explore BrowserAct Skills →

Stop writing automation&scrapers

Install the CLI. Run your first Skill in 30 seconds. Scale when you're ready.

Start free
free · no credit card