Entering Text logo

Entering Text

Community
skydoves
entering-text

Use this skill to drive Jetpack Compose text fields from tests with the text-specific actions — performTextInput (insert at cursor via the InsertTextAtCursor semantics action), performTextReplacement (clear + replace via SetText), performTextClearance (replace with empty string), performTextInputSelection (set TextRange selection), and performImeAction (fire the configured IME action such as Next, Done, Search). Covers the auto-focus path, the enabled / editable / focusable preconditions, the append vs replace gotcha that produces "helloworld" instead of "world", and the IME-action focus chain (Next advances focus to the next focusable). Use when the developer asks "how do I type into a TextField in a test", "test an IME Next button", "test a clear-text action", "BasicTextField won't accept input", "text input duplicated in test", or reports "Failed to perform text input" / "Failed to perform IME action" / "Default ImeAction" errors.

Overview

Publisherskydoves
Repositoryandroid-testing-skills
Skill nameentering-text
Stars
338
Forks
14
Bundled files
Instructions only
LicenseApache-2.0
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.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

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

Installation

Install the Entering Text 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/skydoves/android-testing-skills/tree/main/compose/actions/entering-text
  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/skydoves/android-testing-skills.git /tmp/android-testing-skills
mkdir -p .claude/skills
cp -r /tmp/android-testing-skills/compose/actions/entering-text .claude/skills/entering-text
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Entering Text 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 Entering Text 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 Entering Text 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.

Entering Text — performTextInput, performTextReplacement, performImeAction

Compose text fields expose semantics actions for IME-style edits. The test API wraps those actions so a single call types text, replaces text, clears text, sets a selection, or fires the IME action. Every entry point auto-focuses the node when needed and asserts that the node is enabled and editable, surfacing precise errors when one of those preconditions fails.

When to use this skill

  • The test types into TextField, OutlinedTextField, BasicTextField, or any composable that exposes the InsertTextAtCursor / SetText semantics actions.
  • The test fires an IME action — ImeAction.Next to advance focus, ImeAction.Done to dismiss the keyboard, ImeAction.Search / Send / Go to trigger a form submit.
  • The test must clear an existing value, or replace the full value with new text.
  • The test asserts cursor or selection state with performTextInputSelection(TextRange(...)).

When NOT to use this skill

  • The test sends an arbitrary key shortcut into a focused field (e.g. Ctrl+A, Tab, arrow keys). Use ../injecting-mouse-and-keyboard/SKILL.md.
  • The test only checks the displayed text without typing — assert via ../../assertions/asserting-node-state-and-text/SKILL.md.
  • A finder cannot locate the text field. Verify the matcher first via ../../finders/finding-nodes-by-tag-text-content/SKILL.md.
  • The test asserts state restoration of the typed text after rotation. See ../../setup/configuring-test-dependencies/SKILL.md and StateRestorationTester.

Prerequisites

  • androidx.compose.ui:ui-test-junit4 (or ui-test for runComposeUiTest) configured per ../../setup/configuring-test-dependencies/SKILL.md.
  • The TextField carries Modifier.testTag("…") from production code so the test selects it deterministically.
  • The field is enabled = true. performTextInput calls assert(isEnabled()) before doing anything (TextActions.kt:125).
  • For performImeAction: the field declares both ImeAction != ImeAction.Default (via KeyboardOptions(imeAction = …)) and an OnImeAction handler (via KeyboardActions(...) or default keyboard actions). The action assertion checks both (TextActions.kt:99-100).

Workflow

  1. Type text at the cursor — performTextInput(text) (TextActions.kt:37-41). Internally invokes SemanticsActions.InsertTextAtCursor with AnnotatedString(text). Auto-focuses the node first when it is not focused (getNodeAndFocus — TextActions.kt:119-138). Throws AssertionError("Failed to perform text input.") if the node is not enabled, not focusable, or does not advertise InsertTextAtCursor.
kotlin
rule.onNodeWithTag(Tag).performTextInput("Hello ")
  1. Replace the entire text — performTextReplacement(text) (TextActions.kt:81-84). Invokes SemanticsActions.SetText, which clears existing text and inserts new text. Use this when the field already has a value the test wants to overwrite. MUST NOT be used to "type more characters" — see the WRONG/RIGHT pair below.
kotlin
rule.onNodeWithTag(Tag).performTextReplacement("New value")
  1. Clear the text — performTextClearance() (TextActions.kt:28-30). Sugar for performTextReplacement(""). The field must be focusable + have SetText.
kotlin
rule.onNodeWithTag(Tag).performTextClearance()
  1. Set the selection / cursor — performTextInputSelection(selection: TextRange, relativeToOriginalText: Boolean = true) (TextActions.kt:61-72). Maps to SemanticsActions.SetSelection. When relativeToOriginalText = true, indices refer to the original (untransformed) text; false makes them refer to the post-VisualTransformation text.
kotlin
rule.onNodeWithTag(Tag).performTextInput("Hello World")
rule.onNodeWithTag(Tag).performTextInputSelection(TextRange(0, 5)) // selects "Hello"
  1. Fire the IME action — performImeAction() (TextActions.kt:97-117). Pre-conditions, in order:
    • assert(hasPerformImeAction()) — the field declares OnImeAction.
    • assert(!hasImeAction(ImeAction.Default)) — the action is something other than the default.
    • tryPerformAccessibilityChecks().
    • Auto-focus via getNodeAndFocus(requireEditable = false).
    • Invoke OnImeAction and require it returned true; otherwise throw AssertionError("Failed to perform IME action, handler returned false.").
kotlin
rule.onNodeWithTag(InitialFieldTag).performImeAction()
  1. Pair with assertTextEquals to confirm the expected post-state. The merged-tree default applies to text assertions exactly as to finders — see ../../assertions/asserting-node-state-and-text/SKILL.md.

Patterns

Pattern: Append (performTextInput) vs replace (performTextReplacement)

The most common gotcha — the two APIs look interchangeable but compose differently.

kotlin
// WRONG
rule.onNodeWithTag(Tag).performTextInput("hello")
rule.onNodeWithTag(Tag).performTextInput("world")
rule.onNodeWithTag(Tag).assertTextEquals("world")
// WRONG because: performTextInput maps to InsertTextAtCursor, not SetText. The cursor
// remains at the end of "hello" after the first call, so the second call inserts
// "world" there. Final value is "helloworld" — assertion fails.
kotlin
// RIGHT
rule.onNodeWithTag(Tag).performTextInput("hello")
rule.onNodeWithTag(Tag).performTextReplacement("world")
rule.onNodeWithTag(Tag).assertTextEquals("world")

performTextReplacement calls SemanticsActions.SetText (TextActions.kt:83), which clears existing characters before inserting the new text — equivalent to the IME's "set composing region to whole text and replace" path. Use it any time the second call should overwrite, not append.

Pattern: Canonical insert + assert

From BasicTextFieldTest.kt:171-194:

kotlin
@Test
fun textFieldState_textChange_updatesState() {
    val state = TextFieldState("Hello ", TextRange(Int.MAX_VALUE))
    inputMethodInterceptor.setTextFieldTestContent {
        BasicTextField(state = state, modifier = Modifier.fillMaxSize().testTag(Tag))
    }

    rule.onNodeWithTag(Tag).performTextInput("World!")

    rule.runOnIdle { assertThat(state.text.toString()).isEqualTo("Hello World!") }
}

@Test
fun textFieldState_textChange_updatesSemantics() {
    val state = TextFieldState("Hello ", TextRange(Int.MAX_VALUE))
    inputMethodInterceptor.setTextFieldTestContent {
        BasicTextField(state = state, modifier = Modifier.fillMaxSize().testTag(Tag))
    }

    rule.onNodeWithTag(Tag).performTextInput("World!")

    rule.onNodeWithTag(Tag).assertTextEquals("Hello World!")
}

The cursor is positioned at TextRange(Int.MAX_VALUE) (end of text) when constructing TextFieldState. performTextInput("World!") inserts at that cursor — final value "Hello World!". State reads happen inside runOnIdle (skydoves hot take #5).

Pattern: Replace then assert — BasicTextFieldTest.kt:442-466

kotlin
@Test
fun textField_whenStateObjectChanges_restartsInput() {
    val state1 = TextFieldState("Hello")
    val state2 = TextFieldState("World")
    var toggleState by mutableStateOf(true)
    val state by derivedStateOf { if (toggleState) state1 else state2 }
    inputMethodInterceptor.setTextFieldTestContent {
        BasicTextField(
            state = state,
            enabled = true,
            modifier = Modifier.fillMaxSize().testTag(Tag),
        )
    }

    with(rule.onNodeWithTag(Tag)) {
        performTextReplacement("Compose")
        assertTextEquals("Compose")
    }
    toggleState = !toggleState
    with(rule.onNodeWithTag(Tag)) {
        performTextReplacement("Compose2")
        assertTextEquals("Compose2")
    }
    assertThat(state1.text.toString()).isEqualTo("Compose")
    assertThat(state2.text.toString()).isEqualTo("Compose2")
}

performTextReplacement overwrites whatever the field's current value happens to be — making it the right call when state is rotated underneath.

Pattern: Clearing then refilling

kotlin
rule.onNodeWithTag(Tag).performTextClearance()
rule.onNodeWithTag(Tag).performTextInput("Different")
rule.onNodeWithTag(Tag).assertTextEquals("Different")

performTextClearance is sugar for performTextReplacement("") (TextActions.kt:28-30). After it runs, the cursor is at position 0, so a follow-up performTextInput simply inserts at the start.

Pattern: IME-action focus chain (Next advances focus)

From DefaultKeyboardActionsTest.kt:130-149:

kotlin
// Show keyboard.
rule.onNodeWithTag(initialTextField).performClick()
inputMethodInterceptor.assertSessionActive()
keyboardController.show()

// Act.
rule.onNodeWithTag(initialTextField).performImeAction()

// Assert.
when (imeAction) {
    Next -> {
        // Focus moves to the next item.
        assertThat(focusState1).isFalse()
        assertThat(focusState2).isFalse()
        assertThat(focusState3).isTrue()
    }
    Previous -> {
        // Focus moves to the previous item.
        assertThat(focusState1).isTrue()
        assertThat(focusState2).isFalse()
        assertThat(focusState3).isFalse()
    }
    Done -> {
        // No change to focus state.
        assertThat(focusState1).isFalse()
        assertThat(focusState2).isTrue()
        assertThat(focusState3).isFalse()

        // Software keyboard is hidden.
        keyboardController.assertHidden()
    }
}

Production must declare a non-default ImeAction for performImeAction to be valid:

kotlin
BasicTextField(
    state = state,
    keyboardOptions = KeyboardOptions(imeAction = ImeAction.Next),
    modifier = Modifier.testTag(InitialFieldTag),
)

performImeAction asserts !hasImeAction(ImeAction.Default) first (TextActions.kt:100). Forgetting the keyboardOptions parameter yields AssertionError("Failed to perform IME action.").

Pattern: Set selection before replacing a substring

kotlin
rule.onNodeWithTag(Tag).performTextInput("Hello World")
rule.onNodeWithTag(Tag).performTextInputSelection(TextRange(6, 11)) // select "World"
rule.onNodeWithTag(Tag).performTextInput("Compose")              // selection replaced
rule.onNodeWithTag(Tag).assertTextEquals("Hello Compose")

SemanticsActions.SetSelection honours the same path the IME uses, so a follow-up InsertTextAtCursor replaces the active selection with the new text. Pass relativeToOriginalText = false if the field installs a VisualTransformation (e.g. password mask, currency formatter) and the indices were computed against the displayed string.

Pattern: Disabled or non-editable field — clear error

If the field is enabled = false or the matcher resolves to a Text instead of a TextField, performTextInput throws AssertionError("Failed to perform text input."). The chain is:

  1. tryPerformAccessibilityChecks().
  2. assert(isEnabled()).
  3. assert(hasRequestFocusAction()).
  4. assert(hasSetTextAction()) and assert(hasInsertTextAtCursorAction()) (TextActions.kt:127-129).

When the assertion fires, dump the semantics tree via ../../debug/printing-the-semantics-tree/SKILL.md to confirm the matcher is not selecting a static Text.

Mandatory rules

  • MUST use performTextReplacement to overwrite an existing value. MUST NOT chain two performTextInput calls expecting the second to replace the first — they concatenate at the cursor.
  • MUST declare a non-default ImeAction (e.g. KeyboardOptions(imeAction = ImeAction.Next)) before calling performImeAction. The API asserts !hasImeAction(ImeAction.Default) (TextActions.kt:100) and throws otherwise.
  • MUST select the field by Modifier.testTag(...) whose value is a constant in production code — skydoves hot take #1.
  • MUST route post-input state assertions through rule.runOnIdle { … } (skydoves hot take #5). The recomposer needs to settle before reading TextFieldState.text.
  • MUST NOT call Thread.sleep to wait for IME events. Compose's idle observation includes the text-input pipeline; the assertion will be ready when runOnIdle returns.
  • MUST NOT simulate text entry by injecting individual Key events through performKeyInput { pressKey(Key.A) } unless the test specifically exercises the hardware-keyboard path. The semantics actions are the canonical IME path.
  • PREFERRED: keep performTextInputSelection calls explicit about relativeToOriginalText. The default (true) is correct for most cases; flip it only when the field installs a VisualTransformation.

Verification

  • Every test that overwrites field content uses performTextReplacement(...) or performTextClearance() followed by performTextInput(...). No back-to-back performTextInput calls expecting replacement semantics.
  • Every performImeAction call targets a field that declares KeyboardOptions(imeAction = ImeAction.Next | Done | Search | …) in production source.
  • State or text assertions following an input action run inside rule.runOnIdle { … }.
  • No Thread.sleep in the test method.
  • ./gradlew :app:connectedDebugAndroidTest (or :app:testDebugUnitTest for Robolectric) passes for the test under change.

References

Frequently asked questions

What does the Entering Text AI skill do?

Use this skill to drive Jetpack Compose text fields from tests with the text-specific actions — performTextInput (insert at cursor via the InsertTextAtCursor semantics action), performTextReplacement (clear + replace via SetText), performTextClearance (replace with empty string), performTextInputSelection (set TextRange selection), and performImeAction (fire the configured IME action such as Next, Done, Search). Covers the auto-focus path, the enabled / editable / focusable preconditions, the append vs replace gotcha that produces "helloworld" instead of "world", and the IME-action focus ch...

Why use Entering Text on TypingMind?

Because you install it once and use it with any model. Entering Text 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 Entering Text in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/skydoves/android-testing-skills/tree/main/compose/actions/entering-text. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Entering Text?

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 Entering Text?

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

Is the Entering Text AI skill free?

Yes. It is published on GitHub by skydoves under the Apache-2.0 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 👇