agentleFS
SajidisrarSajidisrar/ios-claude-skills/skills/ios-port-to-android/SKILL.md

ios-port-to-android

Skill1 starsChanged 16 days agoMITView on GitHub

Port an existing iOS app (Swift/SwiftUI/UIKit) to a native Android app (Kotlin/Jetpack Compose) — inventory the iOS app, map each feature and Apple service to its Android equivalent, flag what has no direct equivalent, and rebuild in vertical slices with a parity checklist. Use when asked to bring an iOS app to Android, estimate an Android port, or find the Android equivalent of an iOS API or pattern.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Ask, and record the answer in the project's decision log:

    line 16
  2. Distinctive:

    Use this as the starting point and check each row against current Android documentation (developer.android.com) before relying on it — …

    line 40
  3. Distinctive:

    Keep the backend contract unchanged.

    line 81

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Port an iOS app to Android

A port is a re-implementation of behaviour, not a translation of code. Match what the user can do and what the app guarantees (offline behaviour, security, data model); let the UI follow Material Design and Android conventions rather than copying iOS pixels.

Direction matters: the mirror skill ios-port-from-android covers Android → iOS.

0. Decide the approach first

Ask, and record the answer in the project's decision log:

  • Native Kotlin/Compose rewrite (this skill's default).
  • Share code — e.g. Kotlin Multiplatform for domain/data layers. If the team wants this, the port plan changes (shared modules first); confirm before starting.
  • Minimum Android API level, target devices (phones only? tablets, foldables?), and whether the iOS app keeps evolving during the port (if so, freeze a reference commit).

1. Inventory the iOS app

Produce a table before writing any Kotlin. One row per item:

  • Screens and flows — every screen, its states (loading/empty/error/offline), and how you reach it (tab, push, sheet, deep link).
  • Platform services — every Apple framework the app uses (grep import), entitlements, Info.plist usage strings and URL types, background modes, App Groups, extensions (widgets, notification service, share extension).
  • Third-party SDKs — and whether each has an Android SDK.
  • Backend contract — endpoints, auth flow, push payloads. The API should stay the same; differences found here are backend items, not Android workarounds.
  • Data at rest — what's in SwiftData/Core Data, files, UserDefaults, Keychain, and any migration the Android app must never need (it starts fresh).

2. Map each item

Use this as the starting point and check each row against current Android documentation (developer.android.com) before relying on it — versions and recommended libraries move.

iOSAndroidWatch out for
SwiftUI / UIKitJetpack Compose / ViewsCompose recomposition ≠ SwiftUI invalidation; keep state hoisted
NavigationStack, sheets, tabsNavigation Compose, bottom sheets, NavigationBarSystem back and predictive back must work on every screen
@Observable / ObservableObject view modelViewModel + StateFlowViewModel survives rotation; the screen does not
async/await, Task, actors, @MainActorcoroutines, suspend, viewModelScope, Dispatchers.MainStructured cancellation maps well; actors have no direct twin — use a Mutex or a single-threaded dispatcher
CombineKotlin FlowHot vs cold: StateFlow/SharedFlow vs flow {}
URLSession + CodableOkHttp/Retrofit or Ktor + kotlinx.serializationKeep the same JSON field names and error shapes
SwiftData / Core DataRoomWrite migrations from v1; Room enforces schema at compile time
UserDefaultsDataStoreDataStore is async — no synchronous reads at launch
KeychainAndroid Keystore–backed encryptionKeystore keys can be invalidated (e.g. biometric enrolment change); handle it
BackgroundTasksWorkManagerAndroid doesn't guarantee timing either; design for "eventually"
APNsFirebase Cloud MessagingNotification channels are required; runtime notification permission on newer API levels
StoreKit 2Google Play Billing LibraryServer-side verification and acknowledgement rules differ — re-read Play's docs
Universal Links / apple-app-site-associationAndroid App Links / assetlinks.jsonVerified links need the signing-certificate fingerprint
WidgetKitJetpack Glance / App WidgetsUpdate budgets and sizes differ
Live ActivitiesOngoing notifications (newer Android versions add progress-style live updates)No 1:1 equivalent — agree the design
Privacy manifest, App Privacy labelsPlay Data safety form, manifest permissionsSame facts, different form
Info.plist usage stringsRuntime permission requests + rationale UIAsk in context; handle "don't ask again"
scenePhaseActivity/Lifecycle, SavedStateHandleProcess death: restore state the user would lose
String Catalogsstrings.xml, pluralsPlural categories are per-language on both
Dynamic Type, VoiceOversp font scaling, TalkBack, Compose semanticsTest at the largest font scale
XCTest / Swift TestingJUnit, Compose UI testsPort the test cases, not the test code
Swift Package ManagerGradle (version catalogs)—

Licensing: SF Symbols and the SF Pro fonts are licensed for Apple platforms only. Replace them (e.g. Material Symbols, a licensed or open-source font) — don't bundle them on Android.

Anything without a clear equivalent goes on a gap list with a proposed alternative and an owner (design, product or backend). Don't silently drop a feature.

3. Build in vertical slices

  1. Foundation: project setup, DI, networking client, auth/session, error model, theme.
  2. Then one feature at a time, end to end (data → view model → screen → tests), in the order users need them — usually sign-in, then the main list/detail.
  3. For each slice, port the iOS tests' cases as Android tests first, so behaviour parity is checked, not assumed.
  4. Keep the backend contract unchanged. If Android needs something the API lacks, raise it.

4. Parity check

Per feature, compare against the iOS reference build: same data, same states, same errors, same offline behaviour, same analytics events (if any). Record each as matches, differs by design (Android convention — say which), or gap.

Verification

  • The approach (native vs shared code), minimum API level and iOS reference commit are recorded.
  • Inventory covers every screen, Apple framework, entitlement, extension and SDK.
  • Every mapping used was checked against current Android documentation.
  • Every item without an equivalent is on the gap list with an owner; none was dropped.
  • System back, rotation and process death work on every ported screen.
  • No SF Symbols or SF fonts bundled in the Android app.
  • Backend contract unchanged, or differences raised as backend items.
  • Each feature has a parity verdict, and the iOS test cases exist as Android tests.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.