agentleFS
evancaevanca/flutter-ai-rules/skills/patrol-e2e-testing/SKILL.md

patrol-e2e-testing

Skill669 starsChanged 3 months agoMITView on GitHub

Use when writing E2E/integration tests, testing native interactions like permissions or system dialogs, capturing UI regressions, or validating cross-platform behavior (Patrol 4.x).

The file

Show its rules and text

We split this skill into the rules it gives an agent

16 rules. About 0 are common, 8 are distinctive, and 8 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 8

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill when:

    line 13
  2. Distinctive:

    Follow the official Patrol documentation for installation and project initialization: https://patrol.leancode.co/documentation#setup

    line 23
  3. Very specific:

    Add patrol as a dev dependency.

    line 28
  4. Very specific:

    Place tests in patrol_test/.

    line 29
  5. Distinctive:

    Follow these steps when implementing or updating Patrol tests.

    line 35
  6. Distinctive:

    Avoid combining unrelated flows in a single test.

    line 48
  7. Very specific:

    Use patrolTest() instead of testWidgets().

    line 81
  8. Very specific:

    Use $(#keyName) to find widgets by Key.

    line 83
  9. Very specific:

    Use explicit wait conditions (e.g., waitUntilVisible).

    line 84
  10. Distinctive:

    Use native automation only when required by the feature.

    line 104
  11. Distinctive:

    Run a specific file with live reload (development mode):

    line 167
  12. Distinctive:

    Run a specific file:

    line 173
  13. Very specific:

    Never use Future.delayed as a synchronization mechanism.

    line 212
  14. Very specific:

    Use waitUntilVisible or waitUntilExists to wait for UI state.

    line 213
  15. Distinctive:

    When applied, this skill produces:

    line 218
  16. Very specific:

    Checkpoint: Run patrol test --target <file> locally to confirm the test passes before committing.

    line 225

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

Patrol E2E Testing Skill

Design, implement, and run end-to-end (E2E) tests using Patrol 4.x in Flutter projects.

When to Use

Use this skill when:

  • A new screen or user flow needs E2E test coverage.
  • A feature interacts with native components (permissions, notifications, system dialogs, deep links).
  • A UI bug should be captured as a regression test.
  • Cross-platform behavior (Android / iOS / Web) must be validated.
  • Setting up Patrol in a new or existing Flutter project.

Setup

Follow the official Patrol documentation for installation and project initialization: https://patrol.leancode.co/documentation#setup

Key Patrol conventions:

  • Add patrol as a dev dependency.
  • Place tests in patrol_test/.
  • Name test files with a _test.dart suffix.
  • Execute tests with patrol test.

Workflow

Follow these steps when implementing or updating Patrol tests.

1. Identify the user journey

Break the feature into:

  • Actions: taps, scrolls, input, navigation, deep links.
  • Observable outcomes: visible text, screen changes, enabled buttons, dialogs.

Rules:

  • One test per primary (happy-path) journey.
  • Separate tests for critical edge cases.
  • Avoid combining unrelated flows in a single test.

2. Structure the Patrol test

Basic Patrol structure:

import 'package:flutter_test/flutter_test.dart';
import 'package:patrol/patrol.dart';

void main() {
  patrolTest(
    'user can log in successfully',
    ($) async {
      await $.pumpWidgetAndSettle(const MyApp());

      const email = String.fromEnvironment('E2E_EMAIL');
      const password = String.fromEnvironment('E2E_PASSWORD');

      await $(#emailField).enterText(email);
      await $(#passwordField).enterText(password);
      await $(#loginButton).tap();

      await $.waitUntilVisible($(#homeScreenTitle));

      expect($(#homeScreenTitle).text, equals('Welcome'));
    },
  );
}

Key concepts:

  • Use patrolTest() instead of testWidgets().
  • $ is the Patrol tester.
  • Use $(#keyName) to find widgets by Key.
  • Use explicit wait conditions (e.g., waitUntilVisible).

3. Handle native dialogs

For OS-level permission dialogs:

patrolTest('grants camera permission', ($) async {
  await $.pumpWidgetAndSettle(const MyApp());

  await $(#openCameraButton).tap();

  if (await $.native.isPermissionDialogVisible()) {
    await $.native.grantPermission();
  }

  await $.waitUntilVisible($(#cameraPreview));
});

Use native automation only when required by the feature.

4. Selector & interaction quick reference

Finding widgets:

$('some text')        // by text
$(TextField)          // by type
$(Icons.arrow_back)   // by icon

Tapping:

// Tap a widget containing a specific text label
await $(Container).$('click').tap();

// Tap a container that contains an ElevatedButton
await $(Container).containing(ElevatedButton).tap();

// Tap only the enabled ElevatedButton
await $(ElevatedButton)
    .which<ElevatedButton>(
      (b) => b.enabled,
    )
    .tap();

Entering text:

// Enter text into the second TextField on screen
await $(TextField).at(1).enterText('your input');

Scrolling:

await $(widget_you_want_to_scroll_to).scrollTo();

Native interactions:

// Grant permission while app is in use
await $.native.grantPermissionWhenInUse();

// Open notification shade and tap a notification by text
await $.native.openNotifications();
await $.native.tapOnNotificationBySelector(
  Selector(textContains: 'text'),
);

5. Running Patrol tests

Run all tests:

patrol test

Run a specific file with live reload (development mode):

patrol develop -t integration_test/my_test.dart

Run a specific file:

patrol test --target patrol_test/login_test.dart

Run on web:

patrol test --device chrome

Headless web (CI):

patrol test --device chrome --web-headless true

Filter by tags:

patrol test --tags android

6. Stabilization patterns

Flaky tests undermine confidence. Apply these patterns:

// AVOID — arbitrary delay
await Future.delayed(Duration(seconds: 3));

// PREFER — explicit wait condition
await $.waitUntilVisible($(#targetWidget));

// For animations, pump until settled
await $.pumpAndSettle();
  • Never use Future.delayed as a synchronization mechanism.
  • Use waitUntilVisible or waitUntilExists to wait for UI state.
  • Set settleTimeout in PatrolTesterConfig for slow CI environments.

Output requirements

When applied, this skill produces:

  1. Patrol test(s) covering the specified feature.
  2. Any required widget Key additions to production code.
  3. Exact patrol test command(s) to execute locally.
  4. Notes explaining stabilization or timing decisions.

Checkpoint: Run patrol test --target <file> locally to confirm the test passes before committing.

Quality bar

A valid Patrol test must be:

  • Deterministic — no arbitrary delays; uses explicit wait conditions.
  • Readable — clear test name describing the user journey.
  • Minimal but complete — one assertion chain per journey.
  • Secret-safe — credentials loaded from String.fromEnvironment, never hardcoded.
  • CI-ready — passes headless with --web-headless true or on emulator.

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.