KMP Compose Multiplatform Skiko VSync Render-Decoupling Standard

--- description: KMP Compose Multiplatform Skiko VSync Render-Decoupling Standard - Enforces strict decoupling between high-frequency concurrent state dispatches and Skiko's native graphics pipeline (Metal/Vulkan). Prevents Main dispatcher saturation, frame-pacing drops below the 60Hz/120Hz VSync deadline, and device thermal throttling in Kotlin Multiplatform UI layers. globs: **/commonMain/**/*.kt, **/iosMain/**/*.kt, **/androidMain/**/*.kt, **/jvmMain/**/*.kt alwaysApply: true verified_by: "agentrules.org" license: "CC-BY-4.0" source: "https://agentrules.org/rules/kmp-compose-skiko-vsync-render-decoupling-standard" --- # KMP Compose Multiplatform Skiko VSync Render-Decoupling Standard > **Severity**: CRITICAL | **Category**: mobile > **Anti-Pattern Warning**: Unconflated Real-Time State Ingestion Saturating Skiko Rendering Loops and Violating VSync Budgets ## System Constraints You are a Principal Mobile Architect enforcing the KMP Compose Multiplatform Skiko VSync Render-Decoupling Standard. Compose Multiplatform relies on Skiko to render via native graphics APIs (Metal on iOS/macOS, Vulkan/OpenGL on Android/Desktop). You must strictly mandate that high-frequency concurrent state dispatches (e.g., sensor inputs, WebSocket streams, telemetry, physics ticks) are conflated and decoupled from the Main dispatcher and native VSync cadence. Invariant violations lead to skipped CADisplayLink/Choreographer frames, stuttering, and severe thermal throttling. ## Strict Guidelines - Engineers MUST ALWAYS buffer high-frequency upstream flows with `conflate()` or `buffer(Channel.CONFLATED)` before exposing them to Compose UI state collectors. - Engineers MUST NOT execute business logic, structural state mutations, or heavy JSON/payload transformations on `Dispatchers.Main` or Skiko's render thread. - Engineers MUST ALWAYS isolate CPU-intensive reducer pipelines to `Dispatchers.Default` and transition to the UI state layer only via stable, conflated `StateFlow` primitives. - Animations and continuous visual progressions driven outside Compose animation APIs MUST synchronize using `withFrameNanos` or `withFrameMillis` via Compose's `MonotonicFrameClock`, NEVER via arbitrary `delay()` loops. - Composables NEVER directly collect unbounded cold flows from background repositories; state MUST be ingested through lifecycle-aware, conflated `StateFlow` collectors. - Components MUST NOT schedule more than one render recomposition pass per VSync tick (16.6ms at 60Hz, 8.33ms at 120Hz ProMotion). ## Bad vs Good Implementations ### Decoupling High-Frequency Stream via Conflation and MonotonicFrameClock #### Bad Antipattern (AI Common Hallucination) > **Why it fails**: Collecting an unbounded, high-frequency stream directly on the UI dispatcher triggers recompositions faster than the display's VSync interval (e.g., 60Hz/120Hz). Additionally, performing aggregate computation inside the collection block starves Skiko's render loop, causing skipped frames, Metal/Vulkan pipeline hitches, and battery drain. ```kotlin // Anti-Pattern: Direct Main-dispatcher collection of raw 100Hz+ streaming telemetry // causes Compose recomposition thrashing and blocks Skiko's Metal/Vulkan frame-pacing loop. @Composable fun TelemetryMonitor(telemetryRepository: TelemetryRepository) { val coroutineScope = rememberCoroutineScope() var telemetryData by remember { mutableStateOf<TelemetryData?>(null) } LaunchedEffect(Unit) { // Unbounded high-frequency flow collected directly on Main thread telemetryRepository.rawSensorStream .collect { rawSample -> // Heavy computational transform directly on UI thread val processed = rawSample.calculateAggregates() telemetryData = processed // Triggers recomposition at >120Hz } } telemetryData?.let { data -> RenderTelemetryGraph(data) } } ``` #### Production Standard (Correct) > **Standard**: Transforms are offloaded to `Dispatchers.Default`, shielding the UI thread. The stream is conflated with `conflate()`, ensuring that Skiko only processes the latest state available at each display VSync boundary instead of queuing backlogged micro-recompositions. ```kotlin // Production Fix: Offload transform to Dispatchers.Default, conflate emissions, and sync with UI VSync class TelemetryViewModel( private val telemetryRepository: TelemetryRepository ) : ViewModel() { val uiState: StateFlow<TelemetryUiState> = telemetryRepository.rawSensorStream .flowOn(Dispatchers.Default) // Execute source operations off-main .conflate() // Drop intermediate states if consumer is rendering .map { rawSample -> rawSample.calculateAggregates() } .flowOn(Dispatchers.Default) // Heavy computational transform off-main .stateIn( scope = viewModelScope, started = SharingStarted.WhileSubscribed(5_000), initialValue = TelemetryUiState.Loading ) } @Composable fun TelemetryMonitor(viewModel: TelemetryViewModel) { val state by viewModel.uiState.collectAsStateWithLifecycle() when (val current = state) { is TelemetryUiState.Data -> RenderTelemetryGraph(current.payload) is TelemetryUiState.Loading -> CircularProgressIndicator() } } ```

Architectural Specification & Invariant Rationale

Architectural Invariant Design: KMP Compose Multiplatform Skiko VSync Render-Decoupling Standard

KMP Compose Multiplatform Skiko VSync Render-Decoupling Standard defines an industrial-grade behavioral standard for mobile engineering. Enforces strict decoupling between high-frequency concurrent state dispatches and Skiko's native graphics pipeline (Metal/Vulkan). Prevents Main dispatcher saturation, frame-pacing drops below the 60Hz/120Hz VSync deadline, and device thermal throttling in Kotlin Multiplatform UI layers. Modern software architectures require strict adherence to deterministic patterns to guarantee high availability, maintainability, and zero security vulnerabilities.

Framework Paradigm Shift & Drift MechanicsLLMs are trained on millions of open-source repositories containing outdated code, legacy idioms, and unpatched antipatterns. When coding assistants generate implementations for this stack, they default to common statistical tokens: Unconflated Real-Time State Ingestion Saturating Skiko Rendering Loops and Violating VSync Budgets Without an explicit system instruction file, developers spend hours debugging silent failures.

Failure Modes & Production Blast Radius

Hazard Analysis: What Breaks Without This Rule

The following failure modes have been cataloged from hundreds of hours of production code reviews where AI assistants produced non-compliant implementations:

Failure Mode	AI Symptom (The Anti-Pattern)	Severity	Production Impact & Remedy
AI Hallucination & Deprecation Drift	Unconflated Real-Time State Ingestion Saturating Skiko Rendering Loops and Violating VSync Budgets	Critical	Violating this industrial standard introduces regression bugs, security risks, and unexpected runtime failures in production. Enforce explicit behavioral system directives across all AI coding assistants in your repository.
Silent Architecture Inconsistency	AI assistants produce mixed paradigms across files, creating fragmented codebase conventions.	Medium	Increases cognitive load during code reviews and leads to maintenance debt across engineering teams. Pin project conventions in .cursor/rules, CLAUDE.md, and AGENTS.md so all team agents follow identical guardrails.

Architectural Directives in Detail

Step-by-Step Production Directives

Each rule constraint below is injected into the AI agent's system prompt to establish deterministic boundaries:

Engineers MUST ALWAYS buffer high-frequency upstream flows with `conflate()` or `buffer(Channel.CONFLATED)` before exposing them to Compose UI state collectors.

Rationale:AI models frequently generate unconstrained code here because statistical language distributions lack project-specific architectural context. Mandating this directive ensures that generated code adheres to verified industrial production requirements.