Files
clide/test/builtin/claude/conversation_view_test.dart
T
jpmschweitzerandClaude Opus 4.8 3b23da052e fold the Deny & simplify denial instead of a red error (T-340)
A user-initiated denial (Deny & simplify) came back as an isError
tool_result and rendered as a prominent expanded-red "Bash · error"
block — pure noise, since the user chose it. It now folds to a muted,
collapsed "denied" card.

Built as a reusable filter rather than string-matching the note: DenyTool
carries a `quiet` flag, the session collects quiet denials' tool_use_ids,
and ConversationView renders any error whose id is in that set folded +
muted. Genuine tool failures (ids not in the set) keep the expanded-red
treatment. Adding future "expected error" cases is just adding ids.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-10 16:29:30 +02:00

831 lines
40 KiB
Dart

/// Widget tests for the native Claude ConversationView + controller
/// (T-137): renders each transcript item kind as a card, and text
/// selects + copies across cards via ClideSelectionArea (the terminal
/// affordance we keep, T-135).
library;
import 'dart:async';
import 'package:clide/builtin/claude/src/activity_cluster.dart';
import 'package:clide/builtin/claude/src/claude_banner.dart';
import 'package:clide/builtin/claude/src/conversation_controller.dart';
import 'package:clide/builtin/claude/src/conversation_view.dart';
import 'package:clide/builtin/claude/src/image_thumbnail.dart';
import 'package:clide/builtin/claude/src/transcript_publisher.dart';
import 'package:clide/builtin/claude/src/transcript_reader.dart';
import 'package:clide/kernel/src/events/message_bus.dart';
import 'package:clide/widgets/widgets.dart';
import 'package:flutter/services.dart';
import 'package:flutter/widgets.dart' show Builder, Image, FileImage, MediaQuery, ValueKey;
import 'package:flutter_test/flutter_test.dart';
import '../../helpers/kernel_fixture.dart';
import '../../helpers/widget_harness.dart';
final _t = DateTime.utc(2026, 1, 1);
UserMessage _user(String text) => UserMessage(uuid: 'u', timestamp: _t, isSidechain: false, text: text);
AssistantTextMessage _asst(String text) => AssistantTextMessage(uuid: 'a', timestamp: _t, isSidechain: false, text: text);
AssistantThinkingMessage _think(String text) => AssistantThinkingMessage(uuid: 't', timestamp: _t, isSidechain: false, thinking: text);
AssistantToolUse _tool(String name, Map<String, dynamic> input) =>
AssistantToolUse(uuid: 'tu', timestamp: _t, isSidechain: false, toolUseId: 'x1', name: name, input: input);
ToolResultMessage _result(String content, {bool isError = false}) =>
ToolResultMessage(uuid: 'r', timestamp: _t, isSidechain: false, toolUseId: 'x1', content: content, isError: isError);
ImageMessage _image(String path, {String? caption}) => ImageMessage(uuid: 'i', timestamp: _t, isSidechain: false, path: path, caption: caption);
class _MockClipboard {
Map<String, dynamic> _data = {'text': null};
Future<Object?> handleMethodCall(MethodCall call) async {
switch (call.method) {
case 'Clipboard.setData':
_data = Map<String, dynamic>.from(call.arguments as Map);
case 'Clipboard.getData':
return _data;
case 'Clipboard.hasStrings':
final t = _data['text'] as String?;
return {'value': t != null && t.isNotEmpty};
}
return null;
}
String? get text => _data['text'] as String?;
}
void main() {
group('ConversationController', () {
test('accumulates items from the stream and notifies', () async {
final ctrl = StreamController<ConversationItem>();
final c = ConversationController(stream: ctrl.stream);
addTearDown(c.dispose);
var notifications = 0;
c.addListener(() => notifications++);
expect(c.isEmpty, isTrue);
ctrl.add(_user('hi'));
ctrl.add(_asst('hello'));
// Wait past the coalescing timer (zero-duration, fires after the
// microtask queue drains).
await Future<void>.delayed(const Duration(milliseconds: 20));
expect(c.items, hasLength(2));
expect(c.items.first, isA<UserMessage>());
// Notifications are coalesced: a burst of items collapses to a
// single notify so the view rebuilds once, not per item.
expect(notifications, 1);
await ctrl.close();
});
test('onDispose is invoked on dispose', () async {
final ctrl = StreamController<ConversationItem>();
var disposed = false;
final c = ConversationController(stream: ctrl.stream, onDispose: () async => disposed = true);
c.dispose();
expect(disposed, isTrue);
await ctrl.close();
});
test('toolUseById indexes AssistantToolUse by toolUseId (T-168)', () async {
final ctrl = StreamController<ConversationItem>();
final c = ConversationController(stream: ctrl.stream);
addTearDown(c.dispose);
ctrl.add(_tool('Bash', {'command': 'ls'}));
await Future<void>.delayed(const Duration(milliseconds: 20));
expect(c.toolUseById['x1'], isNotNull);
expect(c.toolUseById['x1']!.name, 'Bash');
await ctrl.close();
});
test('partial-uuid items upsert in the controller (T-168)', () async {
final ctrl = StreamController<ConversationItem>();
final c = ConversationController(stream: ctrl.stream);
addTearDown(c.dispose);
// Two partials with the same `partial-` uuid — second replaces first.
ctrl.add(AssistantTextMessage(uuid: 'partial-m1', timestamp: _t, isSidechain: false, text: 'hello'));
await Future<void>.delayed(const Duration(milliseconds: 20));
ctrl.add(AssistantTextMessage(uuid: 'partial-m1', timestamp: _t, isSidechain: false, text: 'hello world'));
await Future<void>.delayed(const Duration(milliseconds: 20));
expect(c.items.whereType<AssistantTextMessage>(), hasLength(1));
expect(c.items.whereType<AssistantTextMessage>().first.text, 'hello world');
await ctrl.close();
});
});
group('ConversationController.fromBus', () {
test('consumes items published on its publisher/channel', () async {
final bus = MessageBus();
addTearDown(bus.dispose);
final c = ConversationController.fromBus(messages: bus);
addTearDown(c.dispose);
bus.publish(ClaudeConversation.publisher, ClaudeConversation.leadChannel, {ClaudeConversation.itemKey: _user('hi')});
bus.publish(ClaudeConversation.publisher, ClaudeConversation.leadChannel, {ClaudeConversation.itemKey: _asst('hello')});
await Future<void>.delayed(const Duration(milliseconds: 20));
expect(c.items, hasLength(2));
expect(c.items.first, isA<UserMessage>());
});
test('ignores other publishers and channels', () async {
final bus = MessageBus();
addTearDown(bus.dispose);
final c = ConversationController.fromBus(messages: bus);
addTearDown(c.dispose);
bus.publish('someone.else', ClaudeConversation.leadChannel, {ClaudeConversation.itemKey: _user('nope')});
bus.publish(ClaudeConversation.publisher, 'conversation/other', {ClaudeConversation.itemKey: _user('nope')});
await Future<void>.delayed(const Duration(milliseconds: 20));
expect(c.items, isEmpty);
});
});
group('ConversationView', () {
late KernelFixture f;
setUp(() async => f = await KernelFixture.create());
tearDown(() => f.dispose());
Future<ConversationController> pumpWith(WidgetTester tester, List<ConversationItem> items,
{Set<String> hiddenToolUseIds = const {},
Map<String, bool> toolUseOutcomes = const {},
Set<String> quietErrorToolUseIds = const {},
FoldLevel foldLevel = FoldLevel.none}) async {
tester.view.physicalSize = const Size(900, 700);
tester.view.devicePixelRatio = 1.0;
addTearDown(() {
tester.view.resetPhysicalSize();
tester.view.resetDevicePixelRatio();
});
final stream = StreamController<ConversationItem>.broadcast();
final c = ConversationController(stream: stream.stream);
addTearDown(c.dispose);
// Disable animations so an in-flight run's ClideSpinner (a perpetual
// animation) renders static and pumpAndSettle can settle (T-296).
await tester.pumpWidget(harness(
f,
Builder(
builder: (ctx) => MediaQuery(
data: MediaQuery.of(ctx).copyWith(disableAnimations: true),
child: ConversationView(
controller: c,
hiddenToolUseIds: hiddenToolUseIds,
toolUseOutcomes: toolUseOutcomes,
quietErrorToolUseIds: quietErrorToolUseIds,
foldLevel: foldLevel),
),
),
));
for (final it in items) {
stream.add(it);
}
await tester.pumpAndSettle();
return c;
}
testWidgets('empty controller shows the waiting hint', (tester) async {
await pumpWith(tester, const []);
expect(find.text('Waiting for Claude…'), findsOneWidget);
});
testWidgets('unfolded conversation cards carry stable per-item identity keys (T-285)', (tester) async {
tester.view.physicalSize = const Size(900, 800);
tester.view.devicePixelRatio = 1.0;
addTearDown(tester.view.resetPhysicalSize);
addTearDown(tester.view.resetDevicePixelRatio);
final stream = StreamController<ConversationItem>.broadcast();
final c = ConversationController(stream: stream.stream);
addTearDown(c.dispose);
await tester.pumpWidget(harness(
f,
Builder(
// The in-flight Bash collapser shows a live spinner (T-305); disable
// animations so it renders static and pumpAndSettle can settle.
builder: (ctx) => MediaQuery(
data: MediaQuery.of(ctx).copyWith(disableAnimations: true),
child: ConversationView(controller: c, foldLevel: FoldLevel.none),
),
),
));
stream.add(AssistantToolUse(uuid: 'A', timestamp: _t, isSidechain: false, toolUseId: 'A', name: 'Bash', input: const {'command': 'echo a'}));
stream.add(AssistantThinkingMessage(uuid: 'B', timestamp: _t, isSidechain: false, thinking: 'thinking body'));
await tester.pumpAndSettle();
// Each top-level card is keyed by its item uuid so a streaming reshape
// pins card State (collapse/hover) to its logical item, not its position.
expect(find.byKey(const ValueKey('turn.A')), findsOneWidget);
expect(find.byKey(const ValueKey('turn.B')), findsOneWidget);
});
testWidgets('a folded activity cluster carries a stable identity key (T-285)', (tester) async {
tester.view.physicalSize = const Size(900, 800);
tester.view.devicePixelRatio = 1.0;
addTearDown(tester.view.resetPhysicalSize);
addTearDown(tester.view.resetDevicePixelRatio);
final stream = StreamController<ConversationItem>.broadcast();
final c = ConversationController(stream: stream.stream);
addTearDown(c.dispose);
await tester.pumpWidget(harness(
f,
Builder(
builder: (ctx) => MediaQuery(
data: MediaQuery.of(ctx).copyWith(disableAnimations: true),
child: ConversationView(controller: c, foldLevel: FoldLevel.tools),
),
),
));
stream.add(AssistantToolUse(uuid: 'A', timestamp: _t, isSidechain: false, toolUseId: 'A', name: 'Bash', input: const {'command': 'echo a'}));
stream.add(AssistantToolUse(uuid: 'B', timestamp: _t, isSidechain: false, toolUseId: 'B', name: 'Read', input: const {'file_path': '/a'}));
await tester.pumpAndSettle();
// The cluster is keyed by its first item's uuid, so re-folding never
// reattaches the holder's expand State to the wrong cluster by position.
expect(find.byKey(const ValueKey('cluster.A')), findsOneWidget);
});
testWidgets('clicking a bare T-ref in a message opens the tickets reader (T-279)', (tester) async {
Message? opened;
final sub = f.services.messages.subscribe(publisher: 'builtin.tickets', channel: 'selection').listen((m) => opened = m);
addTearDown(sub.cancel);
await pumpWith(tester, [_user('please look at T-281')]);
await tester.tap(find.text('T-281'));
await tester.pumpAndSettle();
expect(opened, isNotNull);
expect(opened!.data['id'], 'T-281');
});
testWidgets('clicking a bare D-ref opens the decisions reader (T-279)', (tester) async {
Message? opened;
final sub = f.services.messages.subscribe(publisher: 'builtin.decisions', channel: 'selection').listen((m) => opened = m);
addTearDown(sub.cancel);
await pumpWith(tester, [_asst('we resume cleanly per D-77 today')]);
await tester.tap(find.text('D-77'));
await tester.pumpAndSettle();
expect(opened, isNotNull);
expect(opened!.data['id'], 'D-77');
});
testWidgets('a pasted-image @path token renders an inline thumbnail (T-236/T-254)', (tester) async {
await pumpWith(tester, [_user('look at this @/tmp/clide/paste-1.png please')]);
expect(find.byType(ImageThumbnail), findsOneWidget);
// The prose around the token still renders.
expect(find.textContaining('look at this'), findsWidgets);
});
testWidgets('multiple image tokens each render a thumbnail (T-236)', (tester) async {
await pumpWith(tester, [_user('@/tmp/a.png and @/tmp/b.jpg')]);
expect(find.byType(ImageThumbnail), findsNWidgets(2));
});
testWidgets('a non-image @path stays literal text (T-236)', (tester) async {
await pumpWith(tester, [_user('see @/tmp/notes.txt for details')]);
expect(find.byType(ImageThumbnail), findsNothing);
});
testWidgets('consecutive same-file edits collapse into one "# edits" card (T-296)', (tester) async {
AssistantToolUse edit(String id, String path) =>
AssistantToolUse(uuid: id, timestamp: _t, isSidechain: false, toolUseId: id, name: 'Edit', input: {'file_path': path});
ToolResultMessage ok(String id) => ToolResultMessage(uuid: 'r$id', timestamp: _t, isSidechain: false, toolUseId: id, content: 'done', isError: false);
await pumpWith(
tester,
[
edit('e1', '/lib/x.dart'),
ok('e1'),
edit('e2', '/lib/x.dart'),
ok('e2'),
],
foldLevel: FoldLevel.tools);
// One bundled card labelled "2 edits" with an aggregate status indicator.
expect(find.text('2 edits'), findsOneWidget);
expect(find.byType(ClideStatusIndicator), findsOneWidget);
});
testWidgets('an edit to a different file is not bundled with the first (T-296)', (tester) async {
AssistantToolUse edit(String id, String path) =>
AssistantToolUse(uuid: id, timestamp: _t, isSidechain: false, toolUseId: id, name: 'Edit', input: {'file_path': path});
await pumpWith(tester, [edit('e1', '/a.dart'), edit('e2', '/b.dart')], foldLevel: FoldLevel.tools);
// Two lone edits, different files → no "edits" bundle.
expect(find.textContaining('edits'), findsNothing);
});
testWidgets('meta items fold into a collapsed activity card; tap expands (T-230)', (tester) async {
await pumpWith(
tester,
[
_tool('Bash', const {'command': 'echo hi'}),
// A second, distinct in-flight tool call (T-262 folds a success
// result into its call card, so two *calls* are what make 2 steps).
AssistantToolUse(uuid: 'tu2', timestamp: _t, isSidechain: false, toolUseId: 'x2', name: 'Read', input: const {'file_path': '/a'}),
],
foldLevel: FoldLevel.tools);
// Collapsed by default: one card with a step count, not the raw rows.
expect(find.text('2 steps'), findsOneWidget);
expect(find.bySemanticsLabel('Activity, 2 steps, collapsed'), findsOneWidget);
// Activating it expands to reveal the folded steps.
await tester.tap(find.bySemanticsLabel('Activity, 2 steps, collapsed'));
await tester.pumpAndSettle();
expect(find.bySemanticsLabel('Activity, 2 steps, expanded'), findsOneWidget);
});
testWidgets('a folded success result is not a separate step — merged into its call (T-262 note D)', (tester) async {
await pumpWith(
tester,
[
_tool('Bash', const {'command': 'echo hi'}),
_result('hi there'),
],
foldLevel: FoldLevel.tools);
// The call + its success result is ONE unit now: 1 step, not 2.
expect(find.text('1 step'), findsOneWidget);
expect(find.text('2 steps'), findsNothing);
// The result is not double-rendered: no standalone result card.
expect(find.text('Bash · result'), findsNothing);
});
testWidgets('an image card renders with the "image" label, the file, and its caption (T-249)', (tester) async {
await pumpWith(tester, [_image('/no/such/file.png', caption: 'before the fix')]);
expect(find.text('image'), findsOneWidget);
expect(find.text('before the fix'), findsOneWidget);
// The image is wired to the resolved file path (display-only, D-78).
final img = tester.widget<Image>(find.byType(Image));
expect((img.image as FileImage).file.path, '/no/such/file.png');
});
testWidgets('inject() drives a new image card into a live view (T-249)', (tester) async {
final c = await pumpWith(tester, [_user('hi')]);
expect(find.text('image'), findsNothing);
c.inject(_image('/no/such/shot.png'));
await tester.pumpAndSettle();
expect(find.text('image'), findsOneWidget);
});
testWidgets('a failed result surfaces first-class, not folded (T-230)', (tester) async {
await pumpWith(
tester,
[
_tool('Bash', const {'command': 'boom'}),
_result('error output', isError: true),
],
foldLevel: FoldLevel.tools);
// The tool call folds (1 step); the error result is sticky → no 2-step card.
expect(find.text('1 step'), findsOneWidget);
expect(find.textContaining('error output'), findsWidgets);
});
testWidgets('empty controller shows the provided emptyState instead', (tester) async {
final stream = StreamController<ConversationItem>.broadcast();
final c = ConversationController(stream: stream.stream);
addTearDown(c.dispose);
await tester.pumpWidget(harness(
f,
ConversationView(
controller: c,
emptyState: const ClideText('CUSTOM EMPTY'),
),
));
expect(find.text('CUSTOM EMPTY'), findsOneWidget);
expect(find.text('Waiting for Claude…'), findsNothing);
});
testWidgets('renders a card per item kind with role/tool labels', (tester) async {
await pumpWith(tester, [
_user('a question'),
_asst('an answer'),
_think('hmm'),
_tool('Bash', {'command': 'ls'}),
_result('ok'),
_result('boom', isError: true),
]);
expect(find.text('you'), findsOneWidget);
expect(find.text('claude'), findsOneWidget);
expect(find.text('thinking'), findsOneWidget);
expect(find.text('Bash'), findsOneWidget);
// T-262: a successful result folds into the tool card (no standalone
// "Bash · result"); a failed result keeps its prominent standalone card.
expect(find.text('Bash · result'), findsNothing);
expect(find.text('Bash · error'), findsOneWidget);
});
testWidgets('AskUserQuestion tool-use and its result are hidden (it shows as a prompt)', (tester) async {
await pumpWith(tester, [
_asst('let me ask'),
AssistantToolUse(uuid: 'au', timestamp: _t, isSidechain: false, toolUseId: 'auq1', name: 'AskUserQuestion', input: const {'questions': []}),
ToolResultMessage(uuid: 'ar', timestamp: _t, isSidechain: false, toolUseId: 'auq1', content: 'answered', isError: false),
_asst('thanks'),
]);
expect(find.text('AskUserQuestion'), findsNothing);
expect(find.text('let me ask'), findsOneWidget);
expect(find.text('thanks'), findsOneWidget);
});
testWidgets('a sidechain prompt folds into its Agent card; never labelled "you" (T-263)', (tester) async {
await pumpWith(tester, [
AssistantToolUse(
uuid: 'agt-msg', timestamp: _t, isSidechain: false, toolUseId: 'task1', name: 'Task', input: const {'description': 'explore the codebase'}),
UserMessage(uuid: 'p1', timestamp: _t, isSidechain: true, text: 'find all the widgets'),
]);
// Never the blue "you", and no standalone block (folded → suppressed).
expect(find.text('you'), findsNothing);
expect(find.text('agent prompt'), findsNothing); // not a standalone card here
expect(find.text('Task'), findsOneWidget);
// Collapsed by default: the prompt is hidden.
expect(find.text('find all the widgets'), findsNothing);
// Expand the Agent collapser → a "prompt" segment reveals the folded prompt.
await tester.tap(find.bySemanticsLabel('Task, 1 step, collapsed'));
await tester.pumpAndSettle();
expect(find.text('prompt'), findsOneWidget); // segment sub-label
expect(find.text('find all the widgets'), findsOneWidget);
});
testWidgets('a stream-json prompt folds via parentToolUseId, not "you" (T-338)', (tester) async {
// The live stream-json wire flags sub-agent prompts with parent_tool_use_id
// (the Task tool-use id) and NO isSidechain/parentUuid — the prompt must
// still fold into the Agent card rather than render as a blue "you" turn.
await pumpWith(tester, [
AssistantToolUse(uuid: 'agt-msg', timestamp: _t, isSidechain: false, toolUseId: 'task1', name: 'Task', input: const {'description': 'explore'}),
UserMessage(uuid: 'sp', timestamp: _t, isSidechain: true, parentToolUseId: 'task1', text: 'find all the widgets'),
]);
expect(find.text('you'), findsNothing);
expect(find.text('agent prompt'), findsNothing);
expect(find.text('Task'), findsOneWidget);
expect(find.text('find all the widgets'), findsNothing); // folded, collapsed
await tester.tap(find.bySemanticsLabel('Task, 1 step, collapsed'));
await tester.pumpAndSettle();
expect(find.text('find all the widgets'), findsOneWidget);
});
testWidgets('an orphan sidechain prompt renders as muted "agent prompt", never "you" (T-263)', (tester) async {
// No Agent tool-use to attach to → stays standalone, but relabelled.
await pumpWith(tester, [
UserMessage(uuid: 'orphan', timestamp: _t, isSidechain: true, text: 'orphaned agent instructions'),
]);
expect(find.text('you'), findsNothing);
expect(find.text('agent prompt'), findsOneWidget);
});
testWidgets('parallel agents: each prompt attaches to its own card via parentUuid (T-263)', (tester) async {
// Document order scrambles the prompts so a nearest-preceding heuristic
// would misattach BOTH to agent B; parentUuid must route them correctly.
await pumpWith(tester, [
AssistantToolUse(uuid: 'mA', timestamp: _t, isSidechain: false, toolUseId: 'tA', name: 'Task', input: const {'description': 'agent A'}),
AssistantToolUse(uuid: 'mB', timestamp: _t, isSidechain: false, toolUseId: 'tB', name: 'Task', input: const {'description': 'agent B'}),
UserMessage(uuid: 'pB', timestamp: _t, isSidechain: true, parentUuid: 'mB', text: 'PROMPT FOR B'),
UserMessage(uuid: 'pA', timestamp: _t, isSidechain: true, parentUuid: 'mA', text: 'PROMPT FOR A'),
]);
// Two collapsed Agent collapsers; the first belongs to card A.
expect(find.text('you'), findsNothing);
await tester.tap(find.bySemanticsLabel('Task, 1 step, collapsed').first);
await tester.pumpAndSettle();
// Only card A is expanded → its prompt (A) shows; B's stays folded away.
// Nearest-preceding would have put A's prompt under B, revealing nothing.
expect(find.text('PROMPT FOR A'), findsOneWidget);
expect(find.text('PROMPT FOR B'), findsNothing);
});
testWidgets('the sub-agent run nests in a holder under its Agent card (T-264)', (tester) async {
await pumpWith(tester, [
AssistantToolUse(uuid: 'mA', timestamp: _t, isSidechain: false, toolUseId: 'tA', name: 'Task', input: const {'description': 'explore'}),
UserMessage(uuid: 'p', timestamp: _t, isSidechain: true, parentUuid: 'mA', text: 'go explore'),
// A sidechain tool call — part of the run, chained off the prompt.
AssistantToolUse(
uuid: 's1', timestamp: _t, isSidechain: true, parentUuid: 'p', toolUseId: 'sb', name: 'Bash', input: const {'command': 'grep widgets'}),
]);
expect(find.text('Task'), findsOneWidget);
// The run is a nested holder titled "agent run", collapsed by default —
// the run's Bash card is not loose in the main chain.
expect(find.bySemanticsLabel('agent run, 1 step, collapsed'), findsOneWidget);
expect(find.text('Bash'), findsNothing); // run card hidden while holder collapsed
await tester.tap(find.bySemanticsLabel('agent run, 1 step, collapsed'));
await tester.pumpAndSettle();
expect(find.text('Bash'), findsOneWidget); // the run's tool card, now nested + visible
});
testWidgets('the returned result is not duplicated when the run is shown (T-264 note E)', (tester) async {
await pumpWith(tester, [
AssistantToolUse(uuid: 'mA', timestamp: _t, isSidechain: false, toolUseId: 'tA', name: 'Task', input: const {'description': 'x'}),
UserMessage(uuid: 'p', timestamp: _t, isSidechain: true, parentUuid: 'mA', text: 'go'),
AssistantTextMessage(uuid: 's1', timestamp: _t, isSidechain: true, parentUuid: 'p', text: 'THE FINAL ANSWER'),
// The Task's returned result (main chain) — equals the run's final prose.
ToolResultMessage(uuid: 'tr', timestamp: _t, isSidechain: false, parentUuid: 'mA', toolUseId: 'tA', content: 'THE FINAL ANSWER', isError: false),
]);
// Expand the Agent collapser (its "result" segment would show here if kept)…
await tester.tap(find.bySemanticsLabel('Task, 1 step, collapsed'));
await tester.pumpAndSettle();
// …and the nested run.
await tester.tap(find.bySemanticsLabel('agent run, 1 step, collapsed'));
await tester.pumpAndSettle();
// The answer appears once (in the run), not also as a folded result segment.
expect(find.textContaining('THE FINAL ANSWER'), findsOneWidget);
expect(find.text('result'), findsNothing); // no result sub-label on the Agent card
});
testWidgets('parallel agents: each run nests under its own card via parentUuid (T-264)', (tester) async {
await pumpWith(tester, [
AssistantToolUse(uuid: 'mA', timestamp: _t, isSidechain: false, toolUseId: 'tA', name: 'Task', input: const {'description': 'A'}),
AssistantToolUse(uuid: 'mB', timestamp: _t, isSidechain: false, toolUseId: 'tB', name: 'Task', input: const {'description': 'B'}),
AssistantToolUse(uuid: 'sA', timestamp: _t, isSidechain: true, parentUuid: 'mA', toolUseId: 'sbA', name: 'Bash', input: const {'command': 'CMD_A'}),
AssistantToolUse(uuid: 'sB', timestamp: _t, isSidechain: true, parentUuid: 'mB', toolUseId: 'sbB', name: 'Bash', input: const {'command': 'CMD_B'}),
]);
// Correct routing → two separate 1-step runs. A nearest-preceding heuristic
// would pool both under agent B as one 2-step run.
expect(find.bySemanticsLabel('agent run, 1 step, collapsed'), findsNWidgets(2));
expect(find.bySemanticsLabel('agent run, 2 steps, collapsed'), findsNothing);
});
testWidgets('a successful sidechain result folds into its run tool card, not a separate step (T-264)', (tester) async {
await pumpWith(tester, [
AssistantToolUse(uuid: 'mA', timestamp: _t, isSidechain: false, toolUseId: 'tA', name: 'Task', input: const {'description': 'x'}),
UserMessage(uuid: 'p', timestamp: _t, isSidechain: true, parentUuid: 'mA', text: 'go'),
AssistantToolUse(uuid: 's1', timestamp: _t, isSidechain: true, parentUuid: 'p', toolUseId: 'sb', name: 'Bash', input: const {'command': 'ls'}),
ToolResultMessage(uuid: 'sr', timestamp: _t, isSidechain: true, parentUuid: 's1', toolUseId: 'sb', content: 'file listing', isError: false),
]);
// The tool call + its success result is ONE step in the run, not two.
expect(find.bySemanticsLabel('agent run, 1 step, collapsed'), findsOneWidget);
expect(find.bySemanticsLabel('agent run, 2 steps, collapsed'), findsNothing);
});
testWidgets('sidechain assistant prose is attributed to "agent", not "claude" (T-265)', (tester) async {
// An orphan sidechain prose (no resolvable Agent) renders inline, still
// attributed to the agent — never the main-thread coral "claude".
await pumpWith(tester, [
AssistantTextMessage(uuid: 's', timestamp: _t, isSidechain: true, text: 'sub-agent says hi'),
]);
expect(find.text('agent'), findsOneWidget);
expect(find.text('claude'), findsNothing);
});
testWidgets('sidechain thinking is attributed to "agent thinking" (T-265)', (tester) async {
await pumpWith(tester, [
AssistantThinkingMessage(uuid: 's', timestamp: _t, isSidechain: true, thinking: 'hmm let me think'),
]);
expect(find.text('agent thinking'), findsOneWidget);
expect(find.text('thinking'), findsNothing);
});
testWidgets('main-thread prose + thinking keep "claude"/"thinking" (T-265 unchanged)', (tester) async {
await pumpWith(tester, [_asst('main says hi'), _think('main thought')]);
expect(find.text('claude'), findsOneWidget);
expect(find.text('thinking'), findsOneWidget);
expect(find.text('agent'), findsNothing);
expect(find.text('agent thinking'), findsNothing);
});
testWidgets('a permission-prompted tool-use is hidden but its result is kept', (tester) async {
await pumpWith(
tester,
[
_tool('Write', {'file_path': '/tmp/x'}),
_result('done')
],
hiddenToolUseIds: {'x1'}, // _tool + _result both use toolUseId 'x1'
);
expect(find.text('Write'), findsNothing); // payload hidden
expect(find.text('done'), findsOneWidget); // result kept
// Label now includes paired tool name (T-168).
expect(find.text('Write · result'), findsOneWidget);
});
testWidgets('a resolved permission tool-use is shown collapsed, not hidden', (tester) async {
await pumpWith(
tester,
[
_tool('Write', {'file_path': '/tmp/x'}),
_result('done')
],
hiddenToolUseIds: {'x1'},
toolUseOutcomes: {'x1': true}, // approved
);
final handle = tester.ensureSemantics();
expect(find.text('Write'), findsOneWidget); // shown (resolved) — collapser label
// T-305: the resolved tool is its own collapser, collapsed by default.
expect(find.bySemanticsLabel('Write, 1 step, collapsed'), findsOneWidget);
// T-262: the resolved card folds its result + carries a success check on
// the collapser (the inner card is hidden while collapsed).
expect(find.byWidgetPredicate((w) => w is ClideStatusIndicator && w.status == ClideRunStatus.success), findsOneWidget);
handle.dispose();
});
testWidgets('an injected user message renders as a muted "context" card, not "you"', (tester) async {
await pumpWith(tester, [
UserMessage(uuid: 'i', timestamp: _t, isSidechain: false, text: 'Base directory for this skill: /x\n\n# pql', injected: true),
_user('a real question'),
]);
expect(find.text('context'), findsOneWidget);
expect(find.text('you'), findsOneWidget); // the real one
});
testWidgets('tool-use body: Bash shows the command in the collapsed summary (T-168)', (tester) async {
await pumpWith(tester, [
_tool('Bash', {'command': 'ls -la'})
]);
// Collapser starts collapsed — the command appears as the echoed summary.
expect(find.text('ls -la'), findsOneWidget);
// Expand the collapser to verify the inner body is a bash code block.
await tester.tap(find.bySemanticsLabel('Bash, 1 step, collapsed'));
await tester.pump();
final blocks = tester.widgetList<ClideCodeBlock>(find.byType(ClideCodeBlock)).toList();
expect(blocks.any((b) => b.language == 'bash' && b.source.contains('ls -la')), isTrue);
});
testWidgets('tool-use body: Read/Grep/LS shows a compact path label (T-168)', (tester) async {
await pumpWith(tester, [
_tool('Read', {'file_path': '/foo/bar.dart'})
]);
// The path label appears (collapsed summary or body).
expect(find.text('/foo/bar.dart'), findsOneWidget);
});
testWidgets('a successful result folds into the tool-use card with a check (T-262)', (tester) async {
final handle = tester.ensureSemantics();
await pumpWith(tester, [
_tool('Read', {'file_path': '/foo/bar.dart'}),
_result('final answer = 42;'),
]);
// No standalone success result card — it's folded into the Read card.
expect(find.text('Read · result'), findsNothing);
// The collapser shows a success check while collapsed.
expect(find.byWidgetPredicate((w) => w is ClideStatusIndicator && w.status == ClideRunStatus.success), findsOneWidget);
// Collapsed by default: neither the call body nor the result is shown yet.
expect(find.text('final answer = 42;'), findsNothing);
// Expand: the call segment, a "result" sub-label, and the folded result
// as a colorized code block (Read → grammar from the .dart path).
await tester.tap(find.bySemanticsLabel('Read, 1 step, collapsed'));
await tester.pumpAndSettle();
expect(find.text('result'), findsOneWidget); // segment sub-label
final blocks = tester.widgetList<ClideCodeBlock>(find.byType(ClideCodeBlock)).toList();
expect(blocks.any((b) => b.language == 'dart' && b.source == 'final answer = 42;'), isTrue);
handle.dispose();
});
testWidgets('Bash result folds as a bash code block; in-flight call has no check (T-262)', (tester) async {
final handle = tester.ensureSemantics();
// In-flight: a call with no result yet → no check, no folded result.
await pumpWith(tester, [
_tool('Bash', {'command': 'ls'})
]);
expect(find.bySemanticsLabel('succeeded'), findsNothing);
expect(find.bySemanticsLabel('failed'), findsNothing);
handle.dispose();
});
testWidgets('Bash success result infers the bash language for the folded block (T-262)', (tester) async {
await pumpWith(tester, [
_tool('Bash', {'command': 'echo hi'}),
_result('hi\nthere'),
]);
await tester.tap(find.bySemanticsLabel('Bash, 1 step, collapsed'));
await tester.pumpAndSettle();
final blocks = tester.widgetList<ClideCodeBlock>(find.byType(ClideCodeBlock)).toList();
expect(blocks.any((b) => b.language == 'bash' && b.source == 'hi\nthere'), isTrue);
});
testWidgets('a failed result is NOT folded: standalone error card + red header mark (T-262)', (tester) async {
final handle = tester.ensureSemantics();
await pumpWith(tester, [
_tool('Bash', {'command': 'boom'}),
_result('permission denied', isError: true),
]);
// The error stays a separate prominent card…
expect(find.text('Bash · error'), findsOneWidget);
expect(find.text('permission denied'), findsOneWidget);
// …and the call collapser carries a red failure mark for symmetry (note C).
expect(find.byWidgetPredicate((w) => w is ClideStatusIndicator && w.status == ClideRunStatus.error), findsOneWidget);
handle.dispose();
});
testWidgets('error result label includes paired tool name (T-168)', (tester) async {
await pumpWith(tester, [
_tool('Bash', {'command': 'cat nonexistent'}),
_result('No such file', isError: true),
]);
expect(find.text('Bash · error'), findsOneWidget);
});
testWidgets('a quiet (user-initiated) denial folds to a muted "denied" card (T-340)', (tester) async {
await pumpWith(
tester,
[
_tool('Bash', {'command': 'rm -rf x'}),
_result('Denied — too complex\nretry simpler', isError: true),
],
quietErrorToolUseIds: {'x1'}, // the toolUseId shared by _tool/_result
);
// Labelled "denied", not the loud "error", and folded by default so the
// body (its second line) is hidden behind a collapsed summary.
expect(find.text('Bash · denied'), findsOneWidget);
expect(find.text('Bash · error'), findsNothing);
expect(find.textContaining('retry simpler'), findsNothing);
});
testWidgets('a genuine error (not in the quiet set) stays expanded red (T-340)', (tester) async {
await pumpWith(tester, [
_tool('Bash', {'command': 'rm -rf x'}),
_result('Denied — too complex\nretry simpler', isError: true),
]);
// Same content, but not flagged quiet → the normal expanded error path.
expect(find.text('Bash · error'), findsOneWidget);
expect(find.text('Bash · denied'), findsNothing);
expect(find.textContaining('retry simpler'), findsOneWidget);
});
testWidgets('result without a paired tool_use uses plain "result" label (T-168)', (tester) async {
// Orphan result (no matching tool_use in the controller).
await pumpWith(tester, [
ToolResultMessage(
uuid: 'r-orphan',
timestamp: _t,
isSidechain: false,
toolUseId: 'unknown-id',
content: 'ok',
isError: false,
),
]);
expect(find.text('result'), findsOneWidget);
});
testWidgets('error result defaults expanded so it is visible (T-168)', (tester) async {
// An error result should show its content without requiring an expand tap.
await pumpWith(tester, [
_tool('Bash', {'command': 'bad'}),
_result('permission denied', isError: true),
]);
// Error content visible without expand.
expect(find.text('permission denied'), findsOneWidget);
});
testWidgets('a one-line tool result renders inline (no collapse caret)', (tester) async {
await pumpWith(tester, [_result('hello-from-spike')]);
expect(find.text('hello-from-spike'), findsOneWidget);
expect(find.byType(ClideIcon), findsNothing); // not collapsible → no caret
});
testWidgets('a multi-line tool result starts collapsed with a first-line summary', (tester) async {
await pumpWith(tester, [_result('first line\nsecond line\nthird line')]);
// Collapsed: caret present, summary (first line) shown, full body hidden.
expect(find.byType(ClideIcon), findsOneWidget);
expect(find.text('first line'), findsOneWidget);
expect(find.text('first line\nsecond line\nthird line'), findsNothing);
await tester.tap(find.byType(ClideIcon));
await tester.pump();
expect(find.text('first line\nsecond line\nthird line'), findsOneWidget);
});
testWidgets('select-all + copy spans multiple cards', (tester) async {
final clipboard = _MockClipboard();
TestDefaultBinaryMessengerBinding.instance.defaultBinaryMessenger.setMockMethodCallHandler(SystemChannels.platform, clipboard.handleMethodCall);
addTearDown(() => TestDefaultBinaryMessengerBinding.instance.defaultBinaryMessenger.setMockMethodCallHandler(SystemChannels.platform, null));
await pumpWith(tester, [_user('question text'), _asst('answer text')]);
// Focus the selection region, select all, copy.
final region = find.byType(ClideSelectionArea);
expect(region, findsOneWidget);
await tester.tap(region);
await tester.pump();
Future<void> keys(LogicalKeyboardKey k) async {
await tester.sendKeyDownEvent(LogicalKeyboardKey.control);
await tester.sendKeyDownEvent(k);
await tester.sendKeyUpEvent(k);
await tester.sendKeyUpEvent(LogicalKeyboardKey.control);
await tester.pump();
}
await keys(LogicalKeyboardKey.keyA);
await keys(LogicalKeyboardKey.keyC);
await tester.pump();
final copied = clipboard.text ?? '';
expect(copied, contains('question text'));
expect(copied, contains('answer text'));
});
});
group('ClaudeBanner', () {
late KernelFixture f;
setUp(() async => f = await KernelFixture.create());
tearDown(() => f.dispose());
testWidgets('shows role, workspace, status, and a hint', (tester) async {
await tester.pumpWidget(harness(
f,
const ClaudeBanner(
role: 'primary',
workspace: '/work/space',
statusLine: 'tmux · clide-claude-x',
),
));
await tester.pump();
expect(find.text('Claude'), findsOneWidget);
expect(find.text('primary'), findsOneWidget);
expect(find.text('/work/space'), findsOneWidget);
expect(find.text('tmux · clide-claude-x'), findsOneWidget);
expect(find.textContaining('Warming up'), findsOneWidget);
});
});
}