Adaptive Layout logo

Adaptive Layout

Community
zakariaf
adaptive-layout

Enforces adapting layout by available CONSTRAINTS/size, never device or platform checks — LayoutBuilder + MediaQuery.sizeOf/paddingOf/viewInsetsOf (not .of) for narrow rebuilds, Material 3 window size classes (compact <600, medium 600-840, expanded 840-1200, large >1200) as the breakpoint vocabulary, navigation affordance chosen by width (NavigationBar → NavigationRail → NavigationDrawer), list-detail single-pane-vs-two-pane, readable max-width via ConstrainedBox, Flexible/Expanded/FractionallySizedBox over fixed widths, SafeArea + display cutouts + keyboard insets, never lock orientation, foldable/hinge awareness via MediaQuery.displayFeatures, and golden-matrix verification across sizes. Use when building responsive or adaptive UI, tablet/desktop/foldable support, master-detail or two-pane screens, a NavigationRail-vs-BottomNav shell, breakpoints, LayoutBuilder, MediaQuery sizing, SafeArea/cutouts, or fixing overflow at large widths.

Overview

Publisherzakariaf
RepositoryFlutter-Skills
Skill nameadaptive-layout
Stars
64
Forks
15
Bundled files
5
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 5 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by zakariaf on GitHub. Read the source before you install it.

Installation

Install the Adaptive Layout AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/zakariaf/Flutter-Skills/tree/main/skills/adaptive-layout
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/zakariaf/Flutter-Skills.git /tmp/Flutter-Skills
mkdir -p .claude/skills
cp -r /tmp/Flutter-Skills/skills/adaptive-layout .claude/skills/adaptive-layout
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Adaptive Layout in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Adaptive Layout on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Adaptive Layout is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Adaptive Layout

Adapt to the space you are given, never to the device you think you are on. A phone in a foldable's front display, a resized desktop window, and a tablet in split-screen all defeat Platform.isX checks — but they all report their real constraints. Branch on width, not hardware.

Read the reference for the task at hand:

  • references/window-size-classes.md — the Material 3 window size-class breakpoints as a shared vocabulary, the WindowSizeClass enum pattern, MediaQuery.sizeOf/paddingOf/viewInsetsOf vs .of, readable max-width, SafeArea and display cutouts, keyboard insets, orientation, foldable hinge awareness.
  • references/list-detail-and-navigation.md — single-pane-navigate vs side-by-side two-pane, choosing the navigation affordance by width and coordinating with the go_router shell, keeping selection state in a Notifier so both panes agree.

Run scripts/check_adaptive.sh before a PR.

Non-negotiable rules

  1. Adapt by constraints/size, never by device or platform. No Platform.isAndroid/Platform.isIOS/kIsWeb to pick a layout. Use LayoutBuilder (local box constraints) or MediaQuery.sizeOf(context) (window size). WHY: a resized window, split-screen, and foldable all break device checks; constraints are always true.

  2. Use the Material 3 window size classes as the breakpoint vocabulary. Compact <600, medium 600–840, expanded 840–1200, large 1200–1600, extra-large ≥1600 (logical px width). WHY: these are STANDARD structural breakpoints (not design tokens); one shared enum keeps every screen's breakpoints identical.

  3. Read the narrowest MediaQuery aspect: sizeOf / paddingOf / viewInsetsOf / viewPaddingOf, not MediaQuery.of(context). WHY: .of subscribes the widget to EVERY MediaQuery change (keyboard, rotation, text scale); the aspect getters rebuild only when that one field changes — see flutter-performance.

  4. Pick the navigation affordance by width, not by a per-screen guess. NavigationBar (compact) → NavigationRail (medium/expanded) → NavigationDrawer (large). Decide once in the shell. WHY: mixing affordances across screens disorients; one width→affordance map is consistent by construction. Coordinate with navigation-and-routing's StatefulShellRoute.

  5. List-detail collapses to one pane on compact and splits on expanded. Compact: the list navigates to a detail route. Expanded: list and detail sit side-by-side; selection is shared state. WHY: two panes on a phone are unreadable; one pane on a desktop wastes 70% of the width.

  6. Never lock orientation (SystemChrome.setPreferredOrientations to force portrait). WHY: it breaks tablets, foldables, and accessibility mounts. Let the size classes handle both orientations.

  7. Constrain readable content with a max width; use Flexible/Expanded/FractionallySizedBox, not fixed pixel widths, for top-level regions. WHY: full-width body text on a wide screen is unreadable (~40–75 chars/line is legible); fixed region widths overflow or leave dead space.

  8. Wrap edge content in SafeArea and respect display cutouts + keyboard insets. Read cutouts/system bars via MediaQuery.paddingOf/viewPaddingOf; the keyboard via MediaQuery.viewInsetsOf. WHY: notches, punch-holes, and the on-screen keyboard occlude content that ignores insets.

  9. Never assume a fixed cell/row height. Large text scale can double a cell's height; let content size itself. WHY: a hardcoded SizedBox(height: 48) clips at 200% text scale — see accessibility-as-code and widget-composition.

  10. Verify across sizes with a golden matrix, not one device. Snapshot each adaptive screen at compact/medium/expanded (+ largest text, LTR/RTL). WHY: adaptive bugs only appear at the breakpoint you did not open — see widget-golden-and-a11y-testing.

Size classes as one shared enum

Define the breakpoints once; every screen reads the same map.

dart
enum WindowSizeClass { compact, medium, expanded, large, extraLarge }

WindowSizeClass windowSizeClassFor(double width) => switch (width) {
      < 600 => WindowSizeClass.compact,
      < 840 => WindowSizeClass.medium,
      < 1200 => WindowSizeClass.expanded,
      < 1600 => WindowSizeClass.large,
      _ => WindowSizeClass.extraLarge,
    };

// In a widget: rebuilds only when the window SIZE changes.
final sizeClass = windowSizeClassFor(MediaQuery.sizeOf(context).width);

Prefer LayoutBuilder when the decision depends on the LOCAL box (a widget inside a split view, a card grid), and MediaQuery.sizeOf when it depends on the whole window (the top-level shell).

dart
// Local constraints drive a card grid's column count.
LayoutBuilder(
  builder: (context, constraints) {
    final columns = constraints.maxWidth ~/ 280; // min card width, structural
    return GridView.count(crossAxisCount: columns.clamp(1, 4), children: cards);
  },
);

Readable width and flexible regions

dart
// Center and cap body text to a legible measure; never full-bleed on desktop.
Center(
  child: ConstrainedBox(
    constraints: const BoxConstraints(maxWidth: 720), // structural readable cap
    child: article,
  ),
);

// Split a wide body proportionally, not with magic fixed widths.
Row(
  children: [
    Expanded(flex: 2, child: primaryPane),
    Expanded(flex: 3, child: secondaryPane),
  ],
);

Insets, cutouts, keyboard, foldables

dart
// Keyboard height: rebuilds only when the keyboard shows/hides.
final keyboard = MediaQuery.viewInsetsOf(context).bottom;

// Notch/system-bar padding without subscribing to text-scale/rotation changes.
final safeTop = MediaQuery.paddingOf(context).top;

// Foldable: a vertical fold/hinge spanning the full height => split around it.
final seam = MediaQuery.displayFeaturesOf(context)
    .where((f) =>
        (f.type == DisplayFeatureType.fold ||
            f.type == DisplayFeatureType.hinge) &&
        f.bounds.top == 0 && // starts at the top edge...
        f.bounds.width < f.bounds.height) // ...taller than wide => vertical
    .firstOrNull; // firstOrNull: import 'package:collection/collection.dart';

Wrap the outermost interactive region in SafeArea; opt specific edges out (bottom: false) when a NavigationBar or scrolling body should reach the edge.

Navigation affordance by width

dart
// One place decides the shell chrome; screens stay affordance-agnostic.
Widget shellFor(WindowSizeClass sc, {required Widget body, required int index}) {
  return switch (sc) {
    WindowSizeClass.compact =>
      Scaffold(body: body, bottomNavigationBar: _NavBar(index: index)),
    WindowSizeClass.medium || WindowSizeClass.expanded =>
      Scaffold(body: Row(children: [_Rail(index: index), Expanded(child: body)])),
    WindowSizeClass.large || WindowSizeClass.extraLarge =>
      Scaffold(body: Row(children: [_Drawer(index: index), Expanded(child: body)])),
  };
}

See examples/adaptive_scaffold.dart for the full shell and examples/list_detail_pane.dart for the two-pane pattern.

Anti-patterns

  • if (Platform.isIOS) ... else ... to choose a layout. Branch on MediaQuery.sizeOf(context).width.
  • MediaQuery.of(context).size.width — subscribes to every MediaQuery change. Use MediaQuery.sizeOf(context).
  • SystemChrome.setPreferredOrientations([DeviceOrientation.portraitUp]) to dodge landscape layout. Support both.
  • Container(width: 375) or SizedBox(width: 320) sizing a top-level pane. Use Expanded/Flexible/FractionallySizedBox/ConstrainedBox.
  • A bool isTablet = MediaQuery.of(context).size.shortestSide > 600 flag threaded everywhere. Compute a WindowSizeClass from width at the shell.
  • Two panes shown on compact width, forcing horizontal scroll. Collapse to one pane and navigate.
  • Full-width paragraphs on desktop with no max-width. Cap the measure.
  • Assuming the keyboard is hidden — content jumps under it. Pad by MediaQuery.viewInsetsOf(context).bottom.

Definition of done

  • No Platform.is*/kIsWeb/dart:io used to select a layout (platform-specific plugins are fine).
  • Breakpoints come from the shared WindowSizeClass helper, not ad-hoc numbers scattered per screen.
  • MediaQuery is read via sizeOf/paddingOf/viewInsetsOf/viewPaddingOf, not .of.
  • Navigation affordance is chosen by width in ONE place (the shell), coordinated with navigation-and-routing.
  • List-detail screens collapse to one pane on compact and split on expanded, sharing selection state.
  • No fixed pixel widths on top-level regions; readable content is max-width capped.
  • Orientation is not locked; edge content uses SafeArea; keyboard insets are respected.
  • Golden matrix covers compact/medium/expanded (+ largest text, RTL).
  • scripts/check_adaptive.sh passes.

Related skills

  • widget-composition — structural layout primitives, computed sizing, edge-to-edge vs SafeArea, never-fixed cell heights.
  • navigation-and-routing — the StatefulShellRoute this shell renders; width picks the chrome, routes pick the branch.
  • flutter-performance — why sizeOf/select beat .of for rebuild scope.
  • accessibility-as-code — large text scale interacts with every adaptive cell.
  • widget-golden-and-a11y-testing — the device × text-scale × RTL golden matrix that proves adaptivity.
  • state-management-riverpod — where shared list-detail selection lives.

References

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Adaptive Layout AI skill do?

Enforces adapting layout by available CONSTRAINTS/size, never device or platform checks — LayoutBuilder + MediaQuery.sizeOf/paddingOf/viewInsetsOf (not .of) for narrow rebuilds, Material 3 window size classes (compact <600, medium 600-840, expanded 840-1200, large >1200) as the breakpoint vocabulary, navigation affordance chosen by width (NavigationBar → NavigationRail → NavigationDrawer), list-detail single-pane-vs-two-pane, readable max-width via ConstrainedBox, Flexible/Expanded/FractionallySizedBox over fixed widths, SafeArea + display cutouts + keyboard insets, never lock orientation,...

Why use Adaptive Layout on TypingMind?

Because you install it once and use it with any model. Adaptive Layout is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Adaptive Layout in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/zakariaf/Flutter-Skills/tree/main/skills/adaptive-layout. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Adaptive Layout?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Adaptive Layout?

As many as you like. As long as a model supports skills, you can use Adaptive Layout with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Adaptive Layout AI skill free?

Yes. It is published on GitHub by zakariaf under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇