diff --git a/app/src/lib/i18n/ar.ts b/app/src/lib/i18n/ar.ts index 941997c19..80c328341 100644 --- a/app/src/lib/i18n/ar.ts +++ b/app/src/lib/i18n/ar.ts @@ -2942,6 +2942,7 @@ const messages: TranslationMap = { 'conversations.taskKanban.moveLeft': 'نقل لليسار', 'conversations.taskKanban.moveRight': 'نقل لليمين', 'conversations.taskKanban.title': 'المهام', + 'conversations.threadTodo.title': 'الخطة', 'conversations.taskKanban.approval.default': 'التقصير', 'conversations.taskKanban.approval.notRequired': 'غير مطلوب', 'conversations.taskKanban.approval.notRequiredBadge': 'عدم الموافقة', diff --git a/app/src/lib/i18n/bn.ts b/app/src/lib/i18n/bn.ts index e09d23741..a54f5cda6 100644 --- a/app/src/lib/i18n/bn.ts +++ b/app/src/lib/i18n/bn.ts @@ -3007,6 +3007,7 @@ const messages: TranslationMap = { 'conversations.taskKanban.moveLeft': 'বামে সরান', 'conversations.taskKanban.moveRight': 'ডানে সরান', 'conversations.taskKanban.title': 'টাস্ক', + 'conversations.threadTodo.title': 'পরিকল্পনা', 'conversations.taskKanban.approval.default': 'ডিফল্ট', 'conversations.taskKanban.approval.notRequired': 'প্রয়োজন নেই', 'conversations.taskKanban.approval.notRequiredBadge': 'অনুমোদন করা হবে না', diff --git a/app/src/lib/i18n/de.ts b/app/src/lib/i18n/de.ts index e01f0fdd8..3acd62571 100644 --- a/app/src/lib/i18n/de.ts +++ b/app/src/lib/i18n/de.ts @@ -3077,6 +3077,7 @@ const messages: TranslationMap = { 'conversations.taskKanban.moveLeft': 'Bewege dich nach links', 'conversations.taskKanban.moveRight': 'Bewege dich nach rechts', 'conversations.taskKanban.title': 'Aufgaben', + 'conversations.threadTodo.title': 'Plan', 'conversations.taskKanban.approval.default': 'Standard', 'conversations.taskKanban.approval.notRequired': 'Nicht erforderlich', 'conversations.taskKanban.approval.notRequiredBadge': 'keine Genehmigung', diff --git a/app/src/lib/i18n/en.ts b/app/src/lib/i18n/en.ts index 99eb1c494..2a2139974 100644 --- a/app/src/lib/i18n/en.ts +++ b/app/src/lib/i18n/en.ts @@ -3536,6 +3536,7 @@ const en: TranslationMap = { 'conversations.taskKanban.moveLeft': 'Move left', 'conversations.taskKanban.moveRight': 'Move right', 'conversations.taskKanban.title': 'Tasks', + 'conversations.threadTodo.title': 'Plan', 'conversations.taskKanban.approval.default': 'Default', 'conversations.taskKanban.approval.notRequired': 'Not required', 'conversations.taskKanban.approval.notRequiredBadge': 'no approval', diff --git a/app/src/lib/i18n/es.ts b/app/src/lib/i18n/es.ts index 2777835c2..8b3519d3f 100644 --- a/app/src/lib/i18n/es.ts +++ b/app/src/lib/i18n/es.ts @@ -3056,6 +3056,7 @@ const messages: TranslationMap = { 'conversations.taskKanban.moveLeft': 'Mover a la izquierda', 'conversations.taskKanban.moveRight': 'Mover a la derecha', 'conversations.taskKanban.title': 'Tareas', + 'conversations.threadTodo.title': 'Plan', 'conversations.taskKanban.approval.default': 'Por defecto', 'conversations.taskKanban.approval.notRequired': 'No requerido', 'conversations.taskKanban.approval.notRequiredBadge': 'sin aprobación', diff --git a/app/src/lib/i18n/fr.ts b/app/src/lib/i18n/fr.ts index fa81de79f..80bc3dafd 100644 --- a/app/src/lib/i18n/fr.ts +++ b/app/src/lib/i18n/fr.ts @@ -3071,6 +3071,7 @@ const messages: TranslationMap = { 'conversations.taskKanban.moveLeft': 'Déplacer à gauche', 'conversations.taskKanban.moveRight': 'Déplacer à droite', 'conversations.taskKanban.title': 'Tâches', + 'conversations.threadTodo.title': 'Plan', 'conversations.taskKanban.approval.default': 'Par défaut', 'conversations.taskKanban.approval.notRequired': 'Non requis', 'conversations.taskKanban.approval.notRequiredBadge': 'aucune approbation', diff --git a/app/src/lib/i18n/hi.ts b/app/src/lib/i18n/hi.ts index b4ef5d90b..d650f39b4 100644 --- a/app/src/lib/i18n/hi.ts +++ b/app/src/lib/i18n/hi.ts @@ -3007,6 +3007,7 @@ const messages: TranslationMap = { 'conversations.taskKanban.moveLeft': 'बाएं ले जाएं', 'conversations.taskKanban.moveRight': 'दाएं ले जाएं', 'conversations.taskKanban.title': 'टास्क', + 'conversations.threadTodo.title': 'योजना', 'conversations.taskKanban.approval.default': 'डिफ़ॉल्ट', 'conversations.taskKanban.approval.notRequired': 'आवश्यकता नहीं', 'conversations.taskKanban.approval.notRequiredBadge': 'कोई अनुमोदन नहीं', diff --git a/app/src/lib/i18n/id.ts b/app/src/lib/i18n/id.ts index a3949c49a..889b1c12e 100644 --- a/app/src/lib/i18n/id.ts +++ b/app/src/lib/i18n/id.ts @@ -3010,6 +3010,7 @@ const messages: TranslationMap = { 'conversations.taskKanban.moveLeft': 'Pindah ke kiri', 'conversations.taskKanban.moveRight': 'Pindah ke kanan', 'conversations.taskKanban.title': 'Tugas', + 'conversations.threadTodo.title': 'Rencana', 'conversations.taskKanban.approval.default': 'Default', 'conversations.taskKanban.approval.notRequired': 'Tidak diperlukan', 'conversations.taskKanban.approval.notRequiredBadge': 'tidak ada persetujuan', diff --git a/app/src/lib/i18n/it.ts b/app/src/lib/i18n/it.ts index c38659517..483bb0279 100644 --- a/app/src/lib/i18n/it.ts +++ b/app/src/lib/i18n/it.ts @@ -3050,6 +3050,7 @@ const messages: TranslationMap = { 'conversations.taskKanban.moveLeft': 'Sposta a sinistra', 'conversations.taskKanban.moveRight': 'Sposta a destra', 'conversations.taskKanban.title': 'Attività', + 'conversations.threadTodo.title': 'Piano', 'conversations.taskKanban.approval.default': 'Predefinito', 'conversations.taskKanban.approval.notRequired': 'Non richiesto', 'conversations.taskKanban.approval.notRequiredBadge': 'nessuna approvazione', diff --git a/app/src/lib/i18n/ko.ts b/app/src/lib/i18n/ko.ts index 03c1e46dd..d54f9593b 100644 --- a/app/src/lib/i18n/ko.ts +++ b/app/src/lib/i18n/ko.ts @@ -2981,6 +2981,7 @@ const messages: TranslationMap = { 'conversations.taskKanban.moveLeft': '왼쪽으로 이동', 'conversations.taskKanban.moveRight': '오른쪽으로 이동', 'conversations.taskKanban.title': '작업', + 'conversations.threadTodo.title': '계획', 'conversations.taskKanban.approval.default': '기본값', 'conversations.taskKanban.approval.notRequired': '필요 없음', 'conversations.taskKanban.approval.notRequiredBadge': '승인 없음', diff --git a/app/src/lib/i18n/pl.ts b/app/src/lib/i18n/pl.ts index fead28d39..3e69c89f6 100644 --- a/app/src/lib/i18n/pl.ts +++ b/app/src/lib/i18n/pl.ts @@ -3037,6 +3037,7 @@ const messages: TranslationMap = { 'conversations.taskKanban.moveLeft': 'Przesuń w lewo', 'conversations.taskKanban.moveRight': 'Przesuń w prawo', 'conversations.taskKanban.title': 'Zadania', + 'conversations.threadTodo.title': 'Plan', 'conversations.taskKanban.approval.default': 'Domyślne', 'conversations.taskKanban.approval.notRequired': 'Niewymagane', 'conversations.taskKanban.approval.notRequiredBadge': 'bez zatwierdzenia', diff --git a/app/src/lib/i18n/pt.ts b/app/src/lib/i18n/pt.ts index 727ef6111..309f0575d 100644 --- a/app/src/lib/i18n/pt.ts +++ b/app/src/lib/i18n/pt.ts @@ -3055,6 +3055,7 @@ const messages: TranslationMap = { 'conversations.taskKanban.moveLeft': 'Mover para esquerda', 'conversations.taskKanban.moveRight': 'Mover para direita', 'conversations.taskKanban.title': 'Tarefas', + 'conversations.threadTodo.title': 'Plano', 'conversations.taskKanban.approval.default': 'Padrão', 'conversations.taskKanban.approval.notRequired': 'Não obrigatório', 'conversations.taskKanban.approval.notRequiredBadge': 'sem aprovação', diff --git a/app/src/lib/i18n/ru.ts b/app/src/lib/i18n/ru.ts index 7b994431e..fbaa9146f 100644 --- a/app/src/lib/i18n/ru.ts +++ b/app/src/lib/i18n/ru.ts @@ -3029,6 +3029,7 @@ const messages: TranslationMap = { 'conversations.taskKanban.moveLeft': 'Переместить влево', 'conversations.taskKanban.moveRight': 'Переместить вправо', 'conversations.taskKanban.title': 'Задачи', + 'conversations.threadTodo.title': 'План', 'conversations.taskKanban.approval.default': 'По умолчанию', 'conversations.taskKanban.approval.notRequired': 'Не требуется', 'conversations.taskKanban.approval.notRequiredBadge': 'нет одобрения', diff --git a/app/src/lib/i18n/zh-CN.ts b/app/src/lib/i18n/zh-CN.ts index 5be6b5093..82e2a4af4 100644 --- a/app/src/lib/i18n/zh-CN.ts +++ b/app/src/lib/i18n/zh-CN.ts @@ -2860,6 +2860,7 @@ const messages: TranslationMap = { 'conversations.taskKanban.moveLeft': '向左移动', 'conversations.taskKanban.moveRight': '向右移动', 'conversations.taskKanban.title': '任务', + 'conversations.threadTodo.title': '计划', 'conversations.taskKanban.approval.default': '默认', 'conversations.taskKanban.approval.notRequired': '无需审批', 'conversations.taskKanban.approval.notRequiredBadge': '无需审批', diff --git a/app/src/pages/Conversations.tsx b/app/src/pages/Conversations.tsx index 9b6851feb..8a1d6f678 100644 --- a/app/src/pages/Conversations.tsx +++ b/app/src/pages/Conversations.tsx @@ -66,7 +66,6 @@ import { } from '../store/threadSlice'; import type { ConfirmationModal as ConfirmationModalType } from '../types/intelligence'; import type { ThreadMessage } from '../types/thread'; -import type { TaskBoardCard, TaskBoardCardStatus } from '../types/turnState'; import { splitAgentMessageIntoBubbles } from '../utils/agentMessageBubbles'; import { chatThreadPath } from '../utils/chatRoutes'; import { CHAT_ATTACHMENTS_ENABLED } from '../utils/config'; @@ -94,7 +93,7 @@ import { } from './conversations/components/BackgroundProcessesPanel'; import { CitationChips, type MessageCitation } from './conversations/components/CitationChips'; import { SubagentDrawer } from './conversations/components/SubagentDrawer'; -import { TaskKanbanBoard } from './conversations/components/TaskKanbanBoard'; +import { ThreadTodoStrip } from './conversations/components/ThreadTodoStrip'; import { ToolTimelineBlock } from './conversations/components/ToolTimelineBlock'; import { evaluateComposerSend, @@ -1412,60 +1411,6 @@ const Conversations = ({ const shouldRenderTimelineBeforeLatestAgentMessage = selectedThreadToolTimeline.length > 0 && !isSending && Boolean(latestVisibleAgentMessage); - const handleMoveTaskCard = async ( - card: TaskBoardCard, - nextStatus: TaskBoardCardStatus - ): Promise => { - if (!selectedThreadId || !selectedTaskBoard) return; - const now = new Date().toISOString(); - const nextBoard = { - ...selectedTaskBoard, - cards: selectedTaskBoard.cards.map(existing => - existing.id === card.id ? { ...existing, status: nextStatus, updatedAt: now } : existing - ), - updatedAt: now, - }; - dispatch(setTaskBoardForThread({ threadId: selectedThreadId, board: nextBoard })); - try { - const saved = await threadApi.putTaskBoard(selectedThreadId, nextBoard.cards); - if (!saved) { - throw new Error('Task board update returned no board'); - } - dispatch(setTaskBoardForThread({ threadId: selectedThreadId, board: saved })); - } catch (error) { - debug('putTaskBoard failed: %o', error); - setSendAdvisory(t('conversations.taskKanban.updateFailed')); - dispatch(setTaskBoardForThread({ threadId: selectedThreadId, board: selectedTaskBoard })); - } - }; - - const handleUpdateTaskCard = async ( - card: TaskBoardCard, - nextCard: TaskBoardCard - ): Promise => { - if (!selectedThreadId || !selectedTaskBoard) return; - const now = new Date().toISOString(); - const nextBoard = { - ...selectedTaskBoard, - cards: selectedTaskBoard.cards.map(existing => - existing.id === card.id ? { ...nextCard, updatedAt: now } : existing - ), - updatedAt: now, - }; - dispatch(setTaskBoardForThread({ threadId: selectedThreadId, board: nextBoard })); - try { - const saved = await threadApi.putTaskBoard(selectedThreadId, nextBoard.cards); - if (!saved) { - throw new Error('Task board update returned no board'); - } - dispatch(setTaskBoardForThread({ threadId: selectedThreadId, board: saved })); - } catch (error) { - debug('putTaskBoard failed: %o', error); - setSendAdvisory(t('conversations.taskKanban.updateFailed')); - dispatch(setTaskBoardForThread({ threadId: selectedThreadId, board: selectedTaskBoard })); - } - }; - const filteredThreads = useMemo(() => { return threads.filter(t => isThreadVisibleInTab(t, selectedLabel)); }, [threads, selectedLabel]); @@ -1805,40 +1750,6 @@ const Conversations = ({ className={`mx-auto w-full max-w-[48.75rem] space-y-3 px-5 pt-4 ${ isSidebar ? 'pb-4' : 'pb-32' }`}> - {selectedTaskBoard && hasTaskBoard && ( - { - void handleMoveTaskCard(card, status); - }} - onUpdateCard={(card, nextCard) => { - void handleUpdateTaskCard(card, nextCard); - }} - onDecidePlan={(card, approve) => { - void runDecidePlan({ - threadId: selectedThreadId, - card, - approve, - dispatch, - notify: setSendAdvisory, - t, - }); - }} - onViewSession={card => { - if (!card.sessionThreadId) return; - // Navigation only — do NOT mark the thread active. activeThreadId - // tracks a true in-flight turn (set on send, cleared on - // done/error). A completed session never emits that lifecycle - // event, so forcing it active would wedge the composer. - dispatch(setSelectedThread(card.sessionThreadId)); - void dispatch(loadThreadMessages(card.sessionThreadId)); - if (shouldSyncChatRoute) { - navigate(chatThreadPath(card.sessionThreadId)); - } - }} - /> - )} {visibleMessages.map(msg => { const isAgentTextMode = msg.sender === 'agent' && agentMessageViewMode === 'text'; // Parsed once per message: for current messages (extraMetadata @@ -2534,6 +2445,38 @@ const Conversations = ({ ); })()} + {/* Thread-scoped todo list the agent maintains as it works — read-only, + pinned above the composer. Distinct from the Intelligence-tab kanban + (global `user-tasks`). Renders nothing when the thread has no active + cards. */} + {selectedThreadId && ( + { + void runDecidePlan({ + threadId: selectedThreadId, + card, + approve, + dispatch, + notify: setSendAdvisory, + t, + }); + }} + onViewSession={card => { + if (!card.sessionThreadId) return; + // Navigation only — do NOT mark the thread active. activeThreadId + // tracks a true in-flight turn; forcing a completed session active + // would wedge the composer. + dispatch(setSelectedThread(card.sessionThreadId)); + void dispatch(loadThreadMessages(card.sessionThreadId)); + if (shouldSyncChatRoute) { + navigate(chatThreadPath(card.sessionThreadId)); + } + }} + /> + )} + {composer === 'mic-cloud' ? (
({ putTaskBoard: vi .fn() .mockResolvedValue({ threadId: 't-1', cards: [], updatedAt: '2026-05-04T10:00:00Z' }), + decidePlan: vi + .fn() + .mockResolvedValue({ threadId: 't-1', cards: [], updatedAt: '2026-05-04T10:00:00Z' }), appendMessage: vi.fn().mockResolvedValue({}), deleteThread: vi.fn().mockResolvedValue({ deleted: true }), generateTitleIfNeeded: vi.fn().mockResolvedValue({}), @@ -1296,109 +1299,6 @@ describe('Conversations — smoke render (#1123 welcome-lock removal)', () => { }); }); - it('rolls back and shows feedback when task board move persistence fails', async () => { - const thread = makeThread({ id: 'board-thread', title: 'Board Thread' }); - const board = { - threadId: 'board-thread', - updatedAt: '2026-05-04T10:00:00Z', - cards: [ - { - id: 'task-1', - title: 'Plan rollout', - status: 'todo' as const, - order: 0, - updatedAt: '2026-05-04T10:00:00Z', - }, - ], - }; - mockGetThreads.mockResolvedValue({ threads: [thread], count: 1 }); - mockGetThreadMessages.mockResolvedValue({ messages: [], count: 0 }); - vi.mocked(threadApi.getTaskBoard).mockResolvedValueOnce(board); - vi.mocked(threadApi.putTaskBoard).mockRejectedValueOnce(new Error('write failed')); - - await act(async () => { - await renderConversations({ - thread: selectedThreadState(thread), - socket: socketState('connected'), - }); - }); - - expect(await screen.findByText('Plan rollout')).toBeInTheDocument(); - fireEvent.click(screen.getByLabelText('Move right')); - - await waitFor(() => { - expect( - screen.getByText('Could not update task; changes were not saved.') - ).toBeInTheDocument(); - }); - // With the 5-column model, todo → right → awaiting_approval (not in_progress) - expect(threadApi.putTaskBoard).toHaveBeenCalledWith( - 'board-thread', - expect.arrayContaining([ - expect.objectContaining({ id: 'task-1', status: 'awaiting_approval' }), - ]) - ); - }); - - it('rolls back and shows feedback when task board edit persistence fails', async () => { - const thread = makeThread({ id: 'edit-board-thread', title: 'Edit Board Thread' }); - const board = { - threadId: 'edit-board-thread', - updatedAt: '2026-05-04T10:00:00Z', - cards: [ - { - id: 'task-1', - title: 'Plan rollout', - status: 'todo' as const, - objective: 'Draft the launch task brief', - assignedAgent: 'planner', - approvalMode: 'required' as const, - plan: ['Read docs'], - allowedTools: ['todo'], - acceptanceCriteria: ['Saved board round-trips'], - evidence: [], - order: 0, - updatedAt: '2026-05-04T10:00:00Z', - }, - ], - }; - mockGetThreads.mockResolvedValue({ threads: [thread], count: 1 }); - mockGetThreadMessages.mockResolvedValue({ messages: [], count: 0 }); - vi.mocked(threadApi.getTaskBoard).mockResolvedValueOnce(board); - vi.mocked(threadApi.putTaskBoard).mockRejectedValueOnce(new Error('write failed')); - - await act(async () => { - await renderConversations({ - thread: selectedThreadState(thread), - socket: socketState('connected'), - }); - }); - - expect(await screen.findByText('Plan rollout')).toBeInTheDocument(); - fireEvent.click(screen.getByText('Task brief')); - fireEvent.change(screen.getByLabelText('Title'), { target: { value: 'Updated rollout' } }); - fireEvent.change(screen.getByLabelText('Assigned agent'), { - target: { value: 'code_executor' }, - }); - fireEvent.click(screen.getByText('Save changes')); - - await waitFor(() => { - expect( - screen.getByText('Could not update task; changes were not saved.') - ).toBeInTheDocument(); - }); - expect(threadApi.putTaskBoard).toHaveBeenCalledWith( - 'edit-board-thread', - expect.arrayContaining([ - expect.objectContaining({ - id: 'task-1', - title: 'Updated rollout', - assignedAgent: 'code_executor', - }), - ]) - ); - }); - it('sends with Enter when the composer is not composing text', async () => { const { textarea, thread } = await renderSelectedConversation(); @@ -2086,4 +1986,41 @@ describe('Conversations — open-session resume (View work)', () => { await waitFor(() => expect(store.getState().thread.selectedThreadId).toBe('sess-99')); expect(screen.getByTestId('route-path')).toHaveTextContent('/human'); }); + + it('approves a parked plan card from the thread todo strip', async () => { + const thread = makeThread({ id: 'approve-thread', title: 'Approve thread' }); + mockGetThreads.mockResolvedValue({ threads: [thread], count: 1 }); + + const store = await renderConversations({ thread: selectedThreadState(thread) }); + const selectedId = store.getState().thread.selectedThreadId ?? 'approve-thread'; + await act(async () => { + store.dispatch( + setTaskBoardForThread({ + threadId: selectedId, + board: { + threadId: selectedId, + updatedAt: '', + cards: [ + { + id: 'pc1', + title: 'Needs sign-off', + status: 'awaiting_approval', + order: 0, + updatedAt: '', + }, + ], + }, + }) + ); + }); + + // The strip surfaces Approve/Reject only for parked cards; approving routes + // through onDecidePlan → runDecidePlan → threadApi.decidePlan. + const approveBtn = await screen.findByTitle('Approve'); + await act(async () => { + fireEvent.click(approveBtn); + }); + + await waitFor(() => expect(threadApi.decidePlan).toHaveBeenCalledWith(selectedId, 'pc1', true)); + }); }); diff --git a/app/src/pages/conversations/components/ThreadTodoStrip.test.tsx b/app/src/pages/conversations/components/ThreadTodoStrip.test.tsx new file mode 100644 index 000000000..a3ad35fcb --- /dev/null +++ b/app/src/pages/conversations/components/ThreadTodoStrip.test.tsx @@ -0,0 +1,176 @@ +import { fireEvent, render, screen } from '@testing-library/react'; +import { describe, expect, it, vi } from 'vitest'; + +import type { TaskBoard, TaskBoardCard } from '../../../types/turnState'; +import { ThreadTodoStrip } from './ThreadTodoStrip'; + +// Echo i18n keys so we can assert on the stable key string. +vi.mock('../../../lib/i18n/I18nContext', () => ({ useT: () => ({ t: (key: string) => key }) })); + +function card(partial: Partial): TaskBoardCard { + return { + id: 'c1', + title: 'Do thing', + status: 'todo', + order: 0, + updatedAt: '', + ...partial, + } as TaskBoardCard; +} + +function board(cards: TaskBoardCard[]): TaskBoard { + return { threadId: 't1', cards, updatedAt: '' }; +} + +describe('ThreadTodoStrip', () => { + it('renders nothing when board is null', () => { + const { container } = render(); + expect(container).toBeEmptyDOMElement(); + }); + + it('renders nothing when there are no active cards', () => { + const { container } = render( + + ); + expect(container).toBeEmptyDOMElement(); + }); + + it('lists active cards and hides done/rejected ones', () => { + render( + + ); + expect(screen.getByText('Active work')).toBeInTheDocument(); + expect(screen.getByText('Queued work')).toBeInTheDocument(); + expect(screen.queryByText('Finished work')).not.toBeInTheDocument(); + expect(screen.queryByText('Dropped work')).not.toBeInTheDocument(); + }); + + it('shows a done/total progress count excluding rejected cards', () => { + render( + + ); + // 1 done out of 2 tracked (rejected excluded entirely). + expect(screen.getByText('1/2')).toBeInTheDocument(); + }); + + it('orders active cards by their `order` field', () => { + render( + + ); + const items = screen.getAllByRole('listitem').map(li => li.textContent); + expect(items[0]).toContain('First'); + expect(items[1]).toContain('Second'); + }); + + it('falls back to objective then id when title is blank', () => { + render( + + ); + expect(screen.getByText('Ship it')).toBeInTheDocument(); + expect(screen.getByText('only-id')).toBeInTheDocument(); + }); + + it('collapses and expands the list when the header is clicked', () => { + render( + + ); + expect(screen.getByText('Active work')).toBeInTheDocument(); + const header = screen.getByRole('button', { expanded: true }); + fireEvent.click(header); + expect(screen.queryByText('Active work')).not.toBeInTheDocument(); + fireEvent.click(screen.getByRole('button', { expanded: false })); + expect(screen.getByText('Active work')).toBeInTheDocument(); + }); + + it('renders Approve/Reject only for awaiting_approval cards when onDecidePlan is provided', () => { + const onDecidePlan = vi.fn(); + render( + + ); + // Exactly one Approve and one Reject — only the parked card has them. + expect(screen.getAllByText('chat.approval.approve')).toHaveLength(1); + expect(screen.getAllByText('chat.approval.deny')).toHaveLength(1); + + fireEvent.click(screen.getByText('chat.approval.approve')); + expect(onDecidePlan).toHaveBeenCalledWith(expect.objectContaining({ id: 'parked' }), true); + fireEvent.click(screen.getByText('chat.approval.deny')); + expect(onDecidePlan).toHaveBeenCalledWith(expect.objectContaining({ id: 'parked' }), false); + }); + + it('surfaces the blocker reason for blocked cards', () => { + render( + + ); + expect(screen.getByText('Stuck step')).toBeInTheDocument(); + expect(screen.getByText('needs API key')).toBeInTheDocument(); + }); + + it('renders a View work jump only for cards with a session when onViewSession is provided', () => { + const onViewSession = vi.fn(); + render( + + ); + expect(screen.getAllByText('conversations.taskKanban.viewWork')).toHaveLength(1); + fireEvent.click(screen.getByText('conversations.taskKanban.viewWork')); + expect(onViewSession).toHaveBeenCalledWith(expect.objectContaining({ id: 'linked' })); + }); + + it('stays fully read-only (no approve/reject) when onDecidePlan is omitted', () => { + render( + + ); + expect(screen.getByText('Needs sign-off')).toBeInTheDocument(); + expect(screen.queryByText('chat.approval.approve')).not.toBeInTheDocument(); + expect(screen.queryByText('chat.approval.deny')).not.toBeInTheDocument(); + }); +}); diff --git a/app/src/pages/conversations/components/ThreadTodoStrip.tsx b/app/src/pages/conversations/components/ThreadTodoStrip.tsx new file mode 100644 index 000000000..4d9592232 --- /dev/null +++ b/app/src/pages/conversations/components/ThreadTodoStrip.tsx @@ -0,0 +1,195 @@ +import React, { useMemo, useState } from 'react'; + +import { useT } from '../../../lib/i18n/I18nContext'; +import type { TaskBoard, TaskBoardCard, TaskBoardCardStatus } from '../../../types/turnState'; + +/** + * Compact, read-only strip that surfaces the *current conversation thread's* + * task board ("todo list") directly above the composer. It mirrors hermes-agent's + * pinned todo panel: a one-glance plan the agent maintains as it works through a + * multi-step task. + * + * Read-only by design — the agent owns the board via its `todo` tool (mutations + * arrive live through the `task_board_updated` socket event into + * `taskBoardByThread`). Direct user editing lives in the full board UI, not here. + * + * This is NOT the Intelligence-tab kanban (that board is the global `user-tasks` + * list). This strip is strictly scoped to the selected chat thread. + */ + +/** Statuses that represent live, actionable work — shown in the strip. */ +const ACTIVE_STATUSES: readonly TaskBoardCardStatus[] = [ + 'in_progress', + 'todo', + 'ready', + 'awaiting_approval', + 'blocked', +]; + +/** Terminal "finished" status — counted toward progress, not listed. */ +const DONE_STATUS: TaskBoardCardStatus = 'done'; + +/** Plain-text glyph per status (monochrome, terminal-style — like hermes). */ +function statusGlyph(status: TaskBoardCardStatus): string { + switch (status) { + case 'in_progress': + return '[~]'; + case 'blocked': + return '[!]'; + case 'awaiting_approval': + return '[?]'; + case 'done': + return '[x]'; + case 'rejected': + return '[-]'; + case 'todo': + case 'ready': + default: + return '[ ]'; + } +} + +/** Tailwind text-color token per status, matching the app's semantic palette. */ +function statusColorClass(status: TaskBoardCardStatus): string { + switch (status) { + case 'in_progress': + return 'text-primary-600 dark:text-primary-300 font-medium'; + case 'blocked': + return 'text-coral dark:text-coral'; + case 'awaiting_approval': + return 'text-amber-700 dark:text-amber-300'; + default: + return 'text-stone-600 dark:text-neutral-300'; + } +} + +interface Props { + board: TaskBoard | null; + /** + * Decide a parked plan (`awaiting_approval` card). When provided, those cards — + * and only those — gain inline Approve/Reject controls; every other card stays + * read-only. Omit to keep the strip fully read-only. + */ + onDecidePlan?: (card: TaskBoardCard, approve: boolean) => void; + /** + * Jump to a card's linked agent session. When provided, cards that carry a + * `sessionThreadId` (stamped by the autonomous/manual task-session flow) gain a + * "View work" affordance. Omit to hide it. + */ + onViewSession?: (card: TaskBoardCard) => void; + /** Disable the approve/reject controls (e.g. no thread selected). */ + disabled?: boolean; +} + +export const ThreadTodoStrip: React.FC = ({ + board, + onDecidePlan, + onViewSession, + disabled = false, +}) => { + const { t } = useT(); + const [collapsed, setCollapsed] = useState(false); + + const { activeCards, doneCount, total } = useMemo(() => { + const cards = board?.cards ?? []; + // Exclude `rejected` cards entirely; they're neither active nor progress. + const tracked = cards.filter(c => c.status !== 'rejected'); + const active = tracked + .filter(c => ACTIVE_STATUSES.includes(c.status)) + .sort((a, b) => a.order - b.order); + return { + activeCards: active, + doneCount: tracked.filter(c => c.status === DONE_STATUS).length, + total: tracked.length, + }; + }, [board]); + + // Nothing to plan → render nothing (no empty chrome above the composer). + if (activeCards.length === 0) return null; + + return ( +
+ + + {!collapsed && ( + // Cap the expanded list so a long plan (one card per step) can't cover + // the latest messages/controls above the composer — scroll instead. +
    + {activeCards.map(card => ( +
  • + + {statusGlyph(card.status)} + + + {cardLabel(card)} + {card.status === 'blocked' && card.blocker?.trim() && ( + // Surface why a step is stuck + what's needed next, matching + // the todo-tool guidance to set `blocked` with a `blocker`. + {card.blocker.trim()} + )} + + {card.sessionThreadId && onViewSession && ( + + )} + {card.status === 'awaiting_approval' && onDecidePlan && ( + + + + + )} +
    • + ))} +
+ )} +
+ ); +}; + +/** Prefer the card title; fall back to its objective, then a generic label. */ +function cardLabel(card: TaskBoardCard): string { + const title = card.title?.trim(); + if (title) return title; + const objective = card.objective?.trim(); + if (objective) return objective; + return card.id; +} + +export default ThreadTodoStrip; diff --git a/app/test/e2e/specs/chat-thread-todo-strip.spec.ts b/app/test/e2e/specs/chat-thread-todo-strip.spec.ts new file mode 100644 index 000000000..269add7d4 --- /dev/null +++ b/app/test/e2e/specs/chat-thread-todo-strip.spec.ts @@ -0,0 +1,128 @@ +// @ts-nocheck +/** + * Chat thread todo strip — end-to-end. + * + * Exercises the per-conversation-thread todo list, top to bottom: + * - The mock LLM emits a `todo` tool call (op:add) on turn 1, then a final + * answer on turn 2. + * - The Rust core executes the thread-bound `todo` tool against the selected + * thread's task board and emits a `task_board_updated` progress event. + * - The frontend records the board into `chatRuntime.taskBoardByThread[tid]`. + * - The read-only `ThreadTodoStrip` (`[data-testid="thread-todo-strip"]`) + * renders above the composer with the agent-authored card. + * + * This is the feature E2E for the "todo list per thread" surface: it proves the + * agent can author its plan AND that the plan renders in the chat view — the + * full path the unit tests can only stub. + */ +import { waitForApp } from '../helpers/app-helpers'; +import { + chatMounted, + clickByTitle, + clickSend, + getSelectedThreadId, + typeIntoComposer, + waitForSocketConnected, +} from '../helpers/chat-harness'; +import { textExists } from '../helpers/element-helpers'; +import { resetApp } from '../helpers/reset-app'; +import { navigateViaHash } from '../helpers/shared-flows'; +import { clearRequestLog, setMockBehavior, startMockServer, stopMockServer } from '../mock-server'; + +const LOG_PREFIX = '[chat-thread-todo-strip]'; +const USER_ID = 'e2e-chat-thread-todo-strip'; +const PROMPT = 'Plan the multi-step refactor before you start.'; +const CARD_TITLE = 'canary-todo-card-7e2a9d'; +const FINAL_REPLY = 'Plan recorded — starting now. canary-final-3f1c'; + +// Turn 1: the LLM writes a card to the thread board via the `todo` tool. +// Turn 2: it answers now that the plan exists. +const FORCED_RESPONSES = [ + { + content: '', + toolCalls: [ + { + id: 'call_todo_add_1', + name: 'todo', + arguments: JSON.stringify({ op: 'add', content: CARD_TITLE, status: 'in_progress' }), + }, + ], + }, + { content: FINAL_REPLY }, +]; + +/** Read the selected thread's board cards straight out of the redux store. */ +async function boardCardTitles(threadId: string): Promise { + return (await browser.execute((tid: string) => { + const winAny = window as unknown as { __OPENHUMAN_STORE__?: { getState: () => unknown } }; + const state = winAny.__OPENHUMAN_STORE__?.getState() as + | { + chatRuntime?: { + taskBoardByThread?: Record }>; + }; + } + | undefined; + const board = state?.chatRuntime?.taskBoardByThread?.[tid]; + return (board?.cards ?? []).map(c => c?.title ?? ''); + }, threadId)) as string[]; +} + +describe('Chat thread todo strip', () => { + before(async function beforeSuite() { + this.timeout(90_000); + await startMockServer(); + await waitForApp(); + await resetApp(USER_ID); + + setMockBehavior('llmForcedResponses', JSON.stringify(FORCED_RESPONSES)); + setMockBehavior('llmStreamChunkDelayMs', '10'); + clearRequestLog(); + console.log(`${LOG_PREFIX} setup complete — forced todo tool-call configured`); + }); + + after(async () => { + setMockBehavior('llmForcedResponses', ''); + setMockBehavior('llmStreamChunkDelayMs', ''); + await stopMockServer(); + }); + + it('renders the agent-authored card in the thread todo strip above the composer', async () => { + await navigateViaHash('/chat'); + await browser.waitUntil(async () => await chatMounted(), { + timeout: 15_000, + timeoutMsg: 'Conversations panel did not mount', + }); + expect(await clickByTitle('New thread', 8_000)).toBe(true); + + const threadId = (await browser.waitUntil(async () => await getSelectedThreadId(), { + timeout: 8_000, + timeoutMsg: 'thread.selectedThreadId never populated', + })) as string; + expect(typeof threadId).toBe('string'); + + await typeIntoComposer(PROMPT); + if (!(await waitForSocketConnected(30_000))) { + console.warn(`${LOG_PREFIX} socket did not connect within 30 s — send may fail`); + } + expect( + await browser.waitUntil(async () => await clickSend(), { + timeout: 5_000, + timeoutMsg: 'Send button never enabled', + }) + ).toBe(true); + + // The board lands in redux once the core processes the `todo` tool call and + // emits task_board_updated. + await browser.waitUntil(async () => (await boardCardTitles(threadId)).includes(CARD_TITLE), { + timeout: 45_000, + timeoutMsg: 'thread board never received the agent-authored card', + }); + + // The read-only strip mounts above the composer and lists the active card. + const strip = await $('[data-testid="thread-todo-strip"]'); + await strip.waitForExist({ timeout: 15_000 }); + expect(await strip.isDisplayed()).toBe(true); + expect(await textExists(CARD_TITLE)).toBe(true); + console.log(`${LOG_PREFIX} passed — strip rendered the agent-authored card`); + }); +}); diff --git a/docs/TEST-COVERAGE-MATRIX.md b/docs/TEST-COVERAGE-MATRIX.md index 8cc00b94f..1366e9f72 100644 --- a/docs/TEST-COVERAGE-MATRIX.md +++ b/docs/TEST-COVERAGE-MATRIX.md @@ -176,12 +176,13 @@ Canonical mapping of every product feature to its test source(s). Drives gap-fil ### 4.2 Messaging -| ID | Feature | Layer | Test path(s) | Status | Notes | -| ----- | -------------------------------------------------------------- | ----- | --------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------- | -| 4.2.1 | User Message Handling | WD+RI | `conversations-web-channel-flow.spec.ts`, `tests/json_rpc_e2e.rs` | ✅ | | -| 4.2.2 | AI Response Generation | WD | `agent-review.spec.ts` | ✅ | Mock LLM | -| 4.2.3 | Streaming Responses | RI | `tests/json_rpc_e2e.rs`, `tests/agent_harness_e2e.rs` | ✅ | `tests/agent_harness_e2e.rs` adds provider-level SSE tool-arg accumulation (chunked args reassembled + parsed) and engine-level delta forwarding (#3471) | -| 4.2.4 | Parallel inference (cross-thread + within-thread forked turns) | RU+VU | `src/openhuman/channels/providers/web_tests.rs`, `app/src/store/__tests__/chatRuntimeSlice.test.ts`, `app/src/providers/__tests__/ChatRuntimeProvider.test.tsx` | 🟡 | Concurrent same-/cross-thread dispatch, cooperative `CancellationToken` teardown, and parallel-lane stream routing covered; dedicated WD E2E is a follow-up | +| ID | Feature | Layer | Test path(s) | Status | Notes | +| ----- | -------------------------------------------------------------- | -------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------- | +| 4.2.1 | User Message Handling | WD+RI | `conversations-web-channel-flow.spec.ts`, `tests/json_rpc_e2e.rs` | ✅ | | +| 4.2.2 | AI Response Generation | WD | `agent-review.spec.ts` | ✅ | Mock LLM | +| 4.2.3 | Streaming Responses | RI | `tests/json_rpc_e2e.rs`, `tests/agent_harness_e2e.rs` | ✅ | `tests/agent_harness_e2e.rs` adds provider-level SSE tool-arg accumulation (chunked args reassembled + parsed) and engine-level delta forwarding (#3471) | +| 4.2.4 | Parallel inference (cross-thread + within-thread forked turns) | RU+VU | `src/openhuman/channels/providers/web_tests.rs`, `app/src/store/__tests__/chatRuntimeSlice.test.ts`, `app/src/providers/__tests__/ChatRuntimeProvider.test.tsx` | 🟡 | Concurrent same-/cross-thread dispatch, cooperative `CancellationToken` teardown, and parallel-lane stream routing covered; dedicated WD E2E is a follow-up | +| 4.2.5 | Per-thread todo list (plan strip above composer) | RU+VU+WD | `src/openhuman/agent/tools/todo.rs`, `app/src/pages/conversations/components/ThreadTodoStrip.test.tsx`, `app/test/e2e/specs/chat-thread-todo-strip.spec.ts` | ✅ | Read-only thread-scoped todo strip fed by `task_board_updated`; agent `todo` tool guidance + thread binding; E2E drives a `todo` tool call and asserts the card renders | ### 4.3 Tool Invocation @@ -199,20 +200,20 @@ Canonical mapping of every product feature to its test source(s). Drives gap-fil End-to-end coverage of the agent harness via the web-chat RPC surface against an in-test scripted upstream (`tests/agent_harness_e2e.rs`) plus browser-level Redux/DOM assertions (`app/test/e2e/specs/agent-harness-behaviors.spec.ts`). Added for #3471. -| ID | Feature | Layer | Test path(s) | Status | Notes | -| ------ | --------------------------------------- | ----- | --------------------------------------------------------------------------------------------------------------------------- | ------ | -------------------------------------------------------------------------------------------------------------- | -| 4.4.1 | Subagent Delegation | RI+WD | `tests/agent_harness_e2e.rs`, `app/test/e2e/specs/agent-harness-behaviors.spec.ts` | ✅ | Orchestrator→researcher delegation asserts ≥3 upstream calls, distinct researcher persona, no "Unknown tool:" | +| ID | Feature | Layer | Test path(s) | Status | Notes | +| ------ | --------------------------------------- | ----- | ---------------------------------------------------------------------------------------------------------------------------- | ------ | --------------------------------------------------------------------------------------------------------------- | +| 4.4.1 | Subagent Delegation | RI+WD | `tests/agent_harness_e2e.rs`, `app/test/e2e/specs/agent-harness-behaviors.spec.ts` | ✅ | Orchestrator→researcher delegation asserts ≥3 upstream calls, distinct researcher persona, no "Unknown tool:" | | 4.4.2 | Subagent Clarification (input-required) | RI+WD | `tests/agent_harness_e2e.rs`, `app/test/e2e/specs/agent-harness-behaviors.spec.ts`, `chat-harness-subagent-continue.spec.ts` | ✅ | Subagent question surfaced + reply completes turn; full `continue_subagent` resume covered by the continue spec | -| 4.4.3 | Approval Gate — Approve | RI+WD | `tests/agent_harness_e2e.rs`, `app/test/e2e/specs/agent-harness-behaviors.spec.ts` | ✅ | `approval_request` event → approve → tool executes (timeline `success`, file written) | -| 4.4.4 | Approval Gate — Deny | RI+WD | `tests/agent_harness_e2e.rs`, `app/test/e2e/specs/agent-harness-behaviors.spec.ts` | ✅ | Deny → tool not executed (timeline `error`, file unchanged), agent explains gracefully | -| 4.4.5 | Approval Gate — TTL Timeout | RI | `tests/agent_harness_e2e.rs` | ✅ | `OPENHUMAN_APPROVAL_TTL_SECS` (debug-only) → no decision → auto-deny | -| 4.4.6 | Subagent + Approval Gate | RI | `tests/agent_harness_e2e.rs` | ✅ | Approval fires inside subagent context; decision propagates; artifact content asserted | -| 4.4.7 | Multi-turn State Persistence | RI | `tests/agent_harness_e2e.rs` | ✅ | Turn-2 upstream request carries turn-1 user + assistant messages | -| 4.4.8 | Parallel Subagent Fan-out | RI | `tests/agent_harness_e2e.rs` | ✅ | `spawn_parallel_agents` 2-task fan-out; both child results reach synthesis call | -| 4.4.9 | Max Iterations / Empty Response | RI | `tests/agent_harness_e2e.rs` | ✅ | Tool-call limit surface + empty-completion graceful terminal (both skip Sentry) | -| 4.4.10 | Provider Error Retry | RI | `tests/agent_harness_e2e.rs` | ✅ | First upstream 500 retried by ReliableProvider; second succeeds | -| 4.4.11 | Inference Phase Transitions | WD | `app/test/e2e/specs/agent-harness-behaviors.spec.ts` | ✅ | Redux `inferenceStatusByThread` observes `subagent` phase then clears to idle | -| 4.4.12 | Tool Timeline Completeness | WD | `app/test/e2e/specs/agent-harness-behaviors.spec.ts` | ✅ | Timeline entries carry id/name/status/round; subagent row reaches `success`; rounds non-decreasing | +| 4.4.3 | Approval Gate — Approve | RI+WD | `tests/agent_harness_e2e.rs`, `app/test/e2e/specs/agent-harness-behaviors.spec.ts` | ✅ | `approval_request` event → approve → tool executes (timeline `success`, file written) | +| 4.4.4 | Approval Gate — Deny | RI+WD | `tests/agent_harness_e2e.rs`, `app/test/e2e/specs/agent-harness-behaviors.spec.ts` | ✅ | Deny → tool not executed (timeline `error`, file unchanged), agent explains gracefully | +| 4.4.5 | Approval Gate — TTL Timeout | RI | `tests/agent_harness_e2e.rs` | ✅ | `OPENHUMAN_APPROVAL_TTL_SECS` (debug-only) → no decision → auto-deny | +| 4.4.6 | Subagent + Approval Gate | RI | `tests/agent_harness_e2e.rs` | ✅ | Approval fires inside subagent context; decision propagates; artifact content asserted | +| 4.4.7 | Multi-turn State Persistence | RI | `tests/agent_harness_e2e.rs` | ✅ | Turn-2 upstream request carries turn-1 user + assistant messages | +| 4.4.8 | Parallel Subagent Fan-out | RI | `tests/agent_harness_e2e.rs` | ✅ | `spawn_parallel_agents` 2-task fan-out; both child results reach synthesis call | +| 4.4.9 | Max Iterations / Empty Response | RI | `tests/agent_harness_e2e.rs` | ✅ | Tool-call limit surface + empty-completion graceful terminal (both skip Sentry) | +| 4.4.10 | Provider Error Retry | RI | `tests/agent_harness_e2e.rs` | ✅ | First upstream 500 retried by ReliableProvider; second succeeds | +| 4.4.11 | Inference Phase Transitions | WD | `app/test/e2e/specs/agent-harness-behaviors.spec.ts` | ✅ | Redux `inferenceStatusByThread` observes `subagent` phase then clears to idle | +| 4.4.12 | Tool Timeline Completeness | WD | `app/test/e2e/specs/agent-harness-behaviors.spec.ts` | ✅ | Timeline entries carry id/name/status/round; subagent row reaches `success`; rounds non-decreasing | --- @@ -274,17 +275,17 @@ End-to-end coverage of the agent harness via the web-chat RPC surface against an ### 6.3 Sub-agent Orchestration -| ID | Feature | Layer | Test path(s) | Status | Notes | -| ----- | ---------------------------- | ----- | --------------------------------------------------------------------------------------------------------------------- | ------ | -------------------------------------------------------------------------------------------------------------------------------------------------- | -| 6.3.1 | Steer a running sub-agent | RU | `src/openhuman/agent_orchestration/running_subagents.rs`, `src/openhuman/agent_orchestration/tools/steer_subagent.rs` | ✅ | `steer_subagent` injects a steer/collect message into a running async sub-agent's run-queue; registry enforces parent ownership + terminal guard. | -| 6.3.2 | Wait for a sub-agent result | RU | `src/openhuman/agent_orchestration/running_subagents.rs`, `src/openhuman/agent_orchestration/tools/wait_subagent.rs` | ✅ | `wait_subagent` blocks on the completion `watch` with a timeout; prunes terminal entries, leaves entries intact on timeout. | -| 6.3.3 | Steer lands in child history | RU | `src/openhuman/agent/harness/subagent_runner/ops_tests.rs::run_queue_steer_lands_in_subagent_history` | ✅ | End-to-end: a queued steer is drained by the child `run_turn_engine` and appears as a `[User steering message]` user turn in the provider request. | -| 6.3.4 | Subconscious trigger pipeline (normalize → dedupe/rate → gate → queue) | RU+RI | `src/openhuman/subconscious_triggers/`, `tests/subconscious_triggers_e2e.rs` | ✅ | Event→Trigger normalization for cron/user/composio/sub-agent, dedupe TTL + per-source rate limit, LLM gate over `agent::triage`, priority queue with overflow eviction. | -| 6.3.5 | Long-lived subconscious orchestrator session | RU | `src/openhuman/subconscious/session.rs`, `src/openhuman/subconscious/user_thread.rs` | ✅ | Persistent compressed session backed by a reserved thread; `notify_user` handoff to the user-facing thread; mode→autonomy config parity. | -| 6.3.6 | Multi-party human↔subconscious↔sub-agent conversation | RI | `tests/subconscious_conversation_e2e.rs` | ✅ | Scripted Gate/SessionExecutor seam drives delegate→sub-agent→merge, failure/retry, interleaving, dedupe, and rate-limit scenarios through the real orchestrator. | -| 6.3.7 | Full-stack trigger pipeline with mocked LLM | RI | `tests/subconscious_fullstack_e2e.rs` (feature `e2e-test-support`) | ✅ | Real `GatePass`+`LongLivedSession`+`Agent`+sub-agent run against a provider-layer mock (no network); promote/drop, persistence, real `spawn_subagent`. | -| 6.3.8 | Subconscious Triggers debug/manage panel (Brain) | WD | `app/test/playwright/specs/subconscious-triggers.spec.ts` | ✅ | Brain→Subconscious panel: renders disabled baseline + hint + reserved thread ids; enable toggle → Pipeline Enabled + event_driven + orchestrator running; disable; refresh re-fetches. | -| 6.3.9 | Vision sub-agent reads attached images | RU | `src/openhuman/agent_registry/agents/loader.rs::vision_agent_loads_on_vision_hint`, `src/openhuman/inference/provider/factory_tests.rs::vision_tier_is_vision_capable`, `src/openhuman/agent/harness/engine/core.rs::gate_tests`, `src/openhuman/agent/multimodal_tests.rs::extract_image_placeholders_pulls_att_tokens_in_order` | ✅ | Orchestrator (non-vision `chat-v1`) keeps the image as a placeholder, delegates to `vision_agent` on the `vision-v1` tier, which rehydrates the on-disk attachment and reads it. Engine gate prefers per-tier `current_model_vision`; turn placeholders forwarded into the sub-agent prompt. | +| ID | Feature | Layer | Test path(s) | Status | Notes | +| ----- | ---------------------------------------------------------------------- | ----- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------ | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | +| 6.3.1 | Steer a running sub-agent | RU | `src/openhuman/agent_orchestration/running_subagents.rs`, `src/openhuman/agent_orchestration/tools/steer_subagent.rs` | ✅ | `steer_subagent` injects a steer/collect message into a running async sub-agent's run-queue; registry enforces parent ownership + terminal guard. | +| 6.3.2 | Wait for a sub-agent result | RU | `src/openhuman/agent_orchestration/running_subagents.rs`, `src/openhuman/agent_orchestration/tools/wait_subagent.rs` | ✅ | `wait_subagent` blocks on the completion `watch` with a timeout; prunes terminal entries, leaves entries intact on timeout. | +| 6.3.3 | Steer lands in child history | RU | `src/openhuman/agent/harness/subagent_runner/ops_tests.rs::run_queue_steer_lands_in_subagent_history` | ✅ | End-to-end: a queued steer is drained by the child `run_turn_engine` and appears as a `[User steering message]` user turn in the provider request. | +| 6.3.4 | Subconscious trigger pipeline (normalize → dedupe/rate → gate → queue) | RU+RI | `src/openhuman/subconscious_triggers/`, `tests/subconscious_triggers_e2e.rs` | ✅ | Event→Trigger normalization for cron/user/composio/sub-agent, dedupe TTL + per-source rate limit, LLM gate over `agent::triage`, priority queue with overflow eviction. | +| 6.3.5 | Long-lived subconscious orchestrator session | RU | `src/openhuman/subconscious/session.rs`, `src/openhuman/subconscious/user_thread.rs` | ✅ | Persistent compressed session backed by a reserved thread; `notify_user` handoff to the user-facing thread; mode→autonomy config parity. | +| 6.3.6 | Multi-party human↔subconscious↔sub-agent conversation | RI | `tests/subconscious_conversation_e2e.rs` | ✅ | Scripted Gate/SessionExecutor seam drives delegate→sub-agent→merge, failure/retry, interleaving, dedupe, and rate-limit scenarios through the real orchestrator. | +| 6.3.7 | Full-stack trigger pipeline with mocked LLM | RI | `tests/subconscious_fullstack_e2e.rs` (feature `e2e-test-support`) | ✅ | Real `GatePass`+`LongLivedSession`+`Agent`+sub-agent run against a provider-layer mock (no network); promote/drop, persistence, real `spawn_subagent`. | +| 6.3.8 | Subconscious Triggers debug/manage panel (Brain) | WD | `app/test/playwright/specs/subconscious-triggers.spec.ts` | ✅ | Brain→Subconscious panel: renders disabled baseline + hint + reserved thread ids; enable toggle → Pipeline Enabled + event_driven + orchestrator running; disable; refresh re-fetches. | +| 6.3.9 | Vision sub-agent reads attached images | RU | `src/openhuman/agent_registry/agents/loader.rs::vision_agent_loads_on_vision_hint`, `src/openhuman/inference/provider/factory_tests.rs::vision_tier_is_vision_capable`, `src/openhuman/agent/harness/engine/core.rs::gate_tests`, `src/openhuman/agent/multimodal_tests.rs::extract_image_placeholders_pulls_att_tokens_in_order` | ✅ | Orchestrator (non-vision `chat-v1`) keeps the image as a placeholder, delegates to `vision_agent` on the `vision-v1` tier, which rehydrates the on-disk attachment and reads it. Engine gate prefers per-tier `current_model_vision`; turn placeholders forwarded into the sub-agent prompt. | --- @@ -349,10 +350,10 @@ End-to-end coverage of the agent harness via the web-chat RPC surface against an ### 8.5 Long-term Goals -| ID | Feature | Test | Source / Test File | Status | Notes | -| ----- | ----------------------------- | ----- | ------------------------------------------------------------------------------------------------------------------------------------------------- | ------ | --------------------------------------------------------------------------------------------- | -| 8.5.1 | Goals CRUD (list/add/edit/delete) | RU+VU | `src/openhuman/memory_goals/store.rs`, `src/openhuman/memory_goals/ops.rs`, `src/openhuman/memory_goals/tools.rs`, `app/src/services/api/goalsApi.test.ts`, `app/src/components/intelligence/GoalsPanel.test.tsx` | ✅ | Editable `MEMORY_GOALS.md` list over `memory_goals_*` RPC + Brain > Goals UI | -| 8.5.2 | Goals enrichment (reflect) | RU+VU | `src/openhuman/memory_goals/enrich.rs`, `src/openhuman/memory_goals/schemas.rs`, `app/src/components/intelligence/GoalsPanel.test.tsx` | 🟡 | Turn-based `goals_agent` enrichment; prompt/registry/error paths unit-tested, live LLM run manual | +| ID | Feature | Test | Source / Test File | Status | Notes | +| ----- | --------------------------------- | ----- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------ | ------------------------------------------------------------------------------------------------- | +| 8.5.1 | Goals CRUD (list/add/edit/delete) | RU+VU | `src/openhuman/memory_goals/store.rs`, `src/openhuman/memory_goals/ops.rs`, `src/openhuman/memory_goals/tools.rs`, `app/src/services/api/goalsApi.test.ts`, `app/src/components/intelligence/GoalsPanel.test.tsx` | ✅ | Editable `MEMORY_GOALS.md` list over `memory_goals_*` RPC + Brain > Goals UI | +| 8.5.2 | Goals enrichment (reflect) | RU+VU | `src/openhuman/memory_goals/enrich.rs`, `src/openhuman/memory_goals/schemas.rs`, `app/src/components/intelligence/GoalsPanel.test.tsx` | 🟡 | Turn-based `goals_agent` enrichment; prompt/registry/error paths unit-tested, live LLM run manual | --- @@ -452,22 +453,22 @@ End-to-end coverage of the agent harness via the web-chat RPC surface against an ### 11.1 Analysis Engine -| ID | Feature | Layer | Test path(s) | Status | Notes | -| ------- | -------------------------------------- | ----- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | ------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------- | -| 11.1.1 | Multi-Source Analysis | RI | `tests/memory_graph_sync_e2e.rs` | 🟡 | Frontend trigger untested | -| 11.1.2 | Actionable Item Extraction | VU | `app/src/components/intelligence/__tests__/utils.test.ts` | ✅ | Was ❌ | -| 11.1.3 | Analyze Trigger | WD | `app/test/e2e/specs/insights-dashboard.spec.ts` mounts the route; explicit analyze-handler invocation TBD | 🟡 | Route mounts and search/filter UI assert — full analyze trigger flow tracked as follow-up | -| 11.1.4 | MCP server (stdio + HTTP) | RU | `src/openhuman/mcp_server/` | ✅ | Stdio framing plus Streamable HTTP/SSE session lifecycle; `McpHttpClient` round-trip tests | -| 11.1.5 | Global tool registry | RI | `src/openhuman/tool_registry/`, `tests/json_rpc_e2e.rs`, `tests/domain_modules_e2e.rs`, `tests/worker_b_domain_e2e.rs` | ✅ | Read-only MCP/controller discovery with routes, schemas, version, allowed agents, and health | -| 11.1.6 | SearXNG MCP search | RU | `src/openhuman/integrations/searxng.rs`, `src/openhuman/mcp_server/tools.rs`, `src/openhuman/tools/schemas.rs` | ✅ | Self-hosted search config, normalized results, MCP argument validation, and mocked HTTP execution | -| 11.1.7 | Bundled prompt resources | RU | `src/openhuman/mcp_server/resources.rs`, `src/openhuman/mcp_server/protocol.rs` | ✅ | `resources/list` catalog + `resources/read` happy path, -32002 unknown URI, -32602 missing param, catalog-mirrors-BUILTINS parity test | -| 11.1.8 | Resource templates list | RU | `src/openhuman/mcp_server/resources.rs`, `src/openhuman/mcp_server/protocol.rs` | ✅ | `resources/templates/list` returns `{resourceTemplates: []}` (static catalog), tolerates unknown/cursor params | -| 11.1.10 | MCP registry install→connect→tool_call | RI | `tests/json_rpc_e2e.rs` (`mcp_clients_install_connect_tool_call_happy_path`), `tests/mcp_registry_e2e.rs`, `src/openhuman/mcp_registry/setup_ops.rs` (#3039) | ✅ | HTTP-RPC happy path install→connect→tool_call→update_env against `test-mcp-stub`; transport-aware install (stdio + http_remote) via `build_install_transport` | -| 11.1.11 | MCP env reconfigure + registry creds | RI/VU | `tests/json_rpc_e2e.rs` (`mcp_clients_registry_settings_roundtrip`), `src/openhuman/mcp_registry/registries/mcp_official.rs`, `app/src/components/channels/mcp/InstalledServerDetail.test.tsx` (#3039) | ✅ | `update_env` persist+reconnect; `registry_settings` get/set with secrets write-only (config-first, env-fallback); reconfigure form validation | -| 11.1.12 | MCP UI surface + setup-agent client | VU | `app/src/components/channels/mcp/InstallDialog.test.tsx`, `app/src/services/api/mcpClientsApi.test.ts`, `app/src/services/api/mcpSetupApi.test.ts` (#3039) | ✅ | Skills `?tab=mcp` renders `McpServersTab` (not Coming Soon); auto-connect on install (best-effort); typed `mcpSetupApi` wrapper | -| 11.1.13 | MCP HTTP-remote auth (token / Bearer / OAuth) + redirect resolution | RU/VU | `src/openhuman/mcp_registry/connections.rs` (`build_http_auth*`, `resolve_final_url`), `src/openhuman/mcp_registry/oauth.rs` (PKCE/token/bundle/callback port), `app/src/components/channels/mcp/ConnectAuthModal.test.tsx` (#3495) | ✅ | Bearer/raw scheme + custom headers; redirect-final-URL resolved before auth; OAuth dynamic client registration + PKCE + refresh; tokens MERGED into stored env; credentials stored encrypted locally, never sent to backend | -| 11.1.14 | MCP "Help & configure" assistant | VU/RU | `app/src/components/channels/mcp/ConfigAssistantPanel.test.tsx`, `app/src/components/channels/mcp/ConfigHelpModal.test.tsx`, `src/openhuman/mcp_registry/ops.rs` (`invoke_config_assist_agent`) (#3495) | ✅ | Fixed server-specific prompt runs an agentic turn scoped to web_search_tool/web_fetch/curl only; markdown-rendered reply; per-MCP chat persisted while on the detail page | -| 11.1.15 | Agent uses connected MCP servers in chat | RU | `src/openhuman/agent_registry/agents/loader.rs` (`orchestrator_subagents_include_mcp_agent`, `mcp_agent_drives_connected_servers_without_install_or_shell`, `planner_has_readonly_mcp_discovery_not_execute`), `src/openhuman/agent_registry/agents/orchestrator/prompt.rs` (`connected_mcp_block_*`), `src/openhuman/agent/harness/session/turn_tests.rs` (`mcp_announcement_fires_once_for_new_server`), `src/openhuman/mcp_registry/{tools,connections}.rs` (#3495) | ✅ | `use_mcp_server` delegate → `mcp_agent` worker (discover→list→call); `mcp_registry_list_tools` read-only discovery; orchestrator `## Connected MCP Servers` prompt block + mid-session connect announcement on the user turn; planner read-only MCP discovery (no `tool_call`) | +| ID | Feature | Layer | Test path(s) | Status | Notes | +| ------- | ------------------------------------------------------------------- | ----- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | +| 11.1.1 | Multi-Source Analysis | RI | `tests/memory_graph_sync_e2e.rs` | 🟡 | Frontend trigger untested | +| 11.1.2 | Actionable Item Extraction | VU | `app/src/components/intelligence/__tests__/utils.test.ts` | ✅ | Was ❌ | +| 11.1.3 | Analyze Trigger | WD | `app/test/e2e/specs/insights-dashboard.spec.ts` mounts the route; explicit analyze-handler invocation TBD | 🟡 | Route mounts and search/filter UI assert — full analyze trigger flow tracked as follow-up | +| 11.1.4 | MCP server (stdio + HTTP) | RU | `src/openhuman/mcp_server/` | ✅ | Stdio framing plus Streamable HTTP/SSE session lifecycle; `McpHttpClient` round-trip tests | +| 11.1.5 | Global tool registry | RI | `src/openhuman/tool_registry/`, `tests/json_rpc_e2e.rs`, `tests/domain_modules_e2e.rs`, `tests/worker_b_domain_e2e.rs` | ✅ | Read-only MCP/controller discovery with routes, schemas, version, allowed agents, and health | +| 11.1.6 | SearXNG MCP search | RU | `src/openhuman/integrations/searxng.rs`, `src/openhuman/mcp_server/tools.rs`, `src/openhuman/tools/schemas.rs` | ✅ | Self-hosted search config, normalized results, MCP argument validation, and mocked HTTP execution | +| 11.1.7 | Bundled prompt resources | RU | `src/openhuman/mcp_server/resources.rs`, `src/openhuman/mcp_server/protocol.rs` | ✅ | `resources/list` catalog + `resources/read` happy path, -32002 unknown URI, -32602 missing param, catalog-mirrors-BUILTINS parity test | +| 11.1.8 | Resource templates list | RU | `src/openhuman/mcp_server/resources.rs`, `src/openhuman/mcp_server/protocol.rs` | ✅ | `resources/templates/list` returns `{resourceTemplates: []}` (static catalog), tolerates unknown/cursor params | +| 11.1.10 | MCP registry install→connect→tool_call | RI | `tests/json_rpc_e2e.rs` (`mcp_clients_install_connect_tool_call_happy_path`), `tests/mcp_registry_e2e.rs`, `src/openhuman/mcp_registry/setup_ops.rs` (#3039) | ✅ | HTTP-RPC happy path install→connect→tool_call→update_env against `test-mcp-stub`; transport-aware install (stdio + http_remote) via `build_install_transport` | +| 11.1.11 | MCP env reconfigure + registry creds | RI/VU | `tests/json_rpc_e2e.rs` (`mcp_clients_registry_settings_roundtrip`), `src/openhuman/mcp_registry/registries/mcp_official.rs`, `app/src/components/channels/mcp/InstalledServerDetail.test.tsx` (#3039) | ✅ | `update_env` persist+reconnect; `registry_settings` get/set with secrets write-only (config-first, env-fallback); reconfigure form validation | +| 11.1.12 | MCP UI surface + setup-agent client | VU | `app/src/components/channels/mcp/InstallDialog.test.tsx`, `app/src/services/api/mcpClientsApi.test.ts`, `app/src/services/api/mcpSetupApi.test.ts` (#3039) | ✅ | Skills `?tab=mcp` renders `McpServersTab` (not Coming Soon); auto-connect on install (best-effort); typed `mcpSetupApi` wrapper | +| 11.1.13 | MCP HTTP-remote auth (token / Bearer / OAuth) + redirect resolution | RU/VU | `src/openhuman/mcp_registry/connections.rs` (`build_http_auth*`, `resolve_final_url`), `src/openhuman/mcp_registry/oauth.rs` (PKCE/token/bundle/callback port), `app/src/components/channels/mcp/ConnectAuthModal.test.tsx` (#3495) | ✅ | Bearer/raw scheme + custom headers; redirect-final-URL resolved before auth; OAuth dynamic client registration + PKCE + refresh; tokens MERGED into stored env; credentials stored encrypted locally, never sent to backend | +| 11.1.14 | MCP "Help & configure" assistant | VU/RU | `app/src/components/channels/mcp/ConfigAssistantPanel.test.tsx`, `app/src/components/channels/mcp/ConfigHelpModal.test.tsx`, `src/openhuman/mcp_registry/ops.rs` (`invoke_config_assist_agent`) (#3495) | ✅ | Fixed server-specific prompt runs an agentic turn scoped to web_search_tool/web_fetch/curl only; markdown-rendered reply; per-MCP chat persisted while on the detail page | +| 11.1.15 | Agent uses connected MCP servers in chat | RU | `src/openhuman/agent_registry/agents/loader.rs` (`orchestrator_subagents_include_mcp_agent`, `mcp_agent_drives_connected_servers_without_install_or_shell`, `planner_has_readonly_mcp_discovery_not_execute`), `src/openhuman/agent_registry/agents/orchestrator/prompt.rs` (`connected_mcp_block_*`), `src/openhuman/agent/harness/session/turn_tests.rs` (`mcp_announcement_fires_once_for_new_server`), `src/openhuman/mcp_registry/{tools,connections}.rs` (#3495) | ✅ | `use_mcp_server` delegate → `mcp_agent` worker (discover→list→call); `mcp_registry_list_tools` read-only discovery; orchestrator `## Connected MCP Servers` prompt block + mid-session connect announcement on the user turn; planner read-only MCP discovery (no `tool_call`) | diff --git a/src/openhuman/agent/tools/todo.rs b/src/openhuman/agent/tools/todo.rs index 85fd7332a..8efe41024 100644 --- a/src/openhuman/agent/tools/todo.rs +++ b/src/openhuman/agent/tools/todo.rs @@ -35,7 +35,17 @@ impl Tool for TodoTool { } fn description(&self) -> &str { - "Manage the agent's per-thread task board. Dispatch via the `op` field: \ + "Maintain a visible plan for THIS conversation thread (the cards render \ + above the user's composer and survive across turns). \ + Use it for any request with 3+ steps or several distinct tasks: at the \ + start, `add` one card per step up front; keep exactly ONE card \ + `in_progress` at a time; mark a card `done` the moment it is finished \ + (do not batch completions); if a step fails or is abandoned, set it \ + `blocked` with a `blocker`, or revise it. `list` to re-read the current \ + plan when resuming. Skip this tool for trivial single-step requests. \ + The board is bound automatically to the current thread — do not pass a \ + thread id. \ + Dispatch via the `op` field: \ `add` (content, status?, objective?, plan?, assignedAgent?, allowedTools?, \ approvalMode?, acceptanceCriteria?, evidence?, notes?, blocker?), \ `edit` (id, content?, status?, objective?, plan?, assignedAgent?, allowedTools?, \ @@ -45,7 +55,6 @@ impl Tool for TodoTool { `replace` (cards: full list — wholesale replace), \ `clear`, or `list`. \ `status` is one of `todo` / `in_progress` / `blocked` / `done`. \ - At most one card may be `in_progress` at a time. \ Returns the updated list as cards plus a markdown rendering." } @@ -347,6 +356,26 @@ mod tests { assert!(err.to_string().contains("content")); } + #[test] + fn description_carries_planning_guidance() { + // The `todo` tool steers the live orchestrator purely through its static + // (prompt-cache-stable) schema description — there is no per-thread prompt + // injection. Lock in the behavioural contract so the guidance can't be + // silently dropped: when-to-use, single-in_progress discipline, and the + // "bound to the current thread, don't pass a thread id" rule. + let tool = TodoTool::new(); + let desc = tool.description(); + assert!(desc.contains("3+ steps"), "missing when-to-use guidance"); + assert!( + desc.contains("ONE card `in_progress`"), + "missing single-in_progress discipline" + ); + assert!( + desc.contains("do not pass a thread id"), + "missing explicit 'do not pass a thread id' note" + ); + } + #[tokio::test] async fn edit_rejects_unknown_id() { let _guard = scratch_lock();