Repository navigation
fix(chat): route OMP question cards to terminal - #31440
Conversation
📝 Walkthrough
Priority: ⚪ Pending latest changes Merge Risk: 🟡 Moderate · up to
|
There was a problem hiding this comment.
Actionable comments posted: 1
ℹ️ Review info
⚙️ Run configuration
- Configuration used: Repository UI
- Review profile: CHILL
- Plan: Advanced
- Run ID:
0fbc6a73-c5f6-4d28-8d59-20d7331bde18
📒 Files selected for processing: 17
mobile/src/session/use-mobile-bridge-chat-prompt-writes.tsmobile/src/session/use-mobile-native-chat-answer-send-omp.test.tsmobile/src/session/use-mobile-native-chat-answer-send.tsmobile/src/session/use-mobile-native-chat-controller.tssrc/renderer/src/components/native-chat/native-chat-input-clear.tssrc/renderer/src/components/native-chat/native-chat-interactive-prompt.test.tssrc/renderer/src/components/native-chat/native-chat-interactive-prompt.tssrc/renderer/src/components/native-chat/native-chat-observed-send.tssrc/renderer/src/components/native-chat/native-chat-runtime-send.test.tssrc/renderer/src/components/native-chat/native-chat-runtime-send.tssrc/renderer/src/components/native-chat/use-native-chat-interactive-send.test.tsxsrc/renderer/src/components/native-chat/use-native-chat-interactive-send.tssrc/shared/native-chat-agent-support.test.tssrc/shared/native-chat-agent-support.tssrc/shared/native-chat-ask.test.tssrc/shared/native-chat-ask.tssrc/shared/native-chat-omp-selection.ts
Included review availability: This review used your included allowance. Your plan provides up to 10 included reviews per hour; 0 remain after this review.
| if (omp && attemptedOmpAskRef.current === baselineKey) { | ||
| onDeliverySettled?.(false) | ||
| return { settleAfterMs: 0 } | ||
| } |
There was a problem hiding this comment.
🎯 Functional Correctness | 🟡 Minor | ⚡ Quick win
🔎 Supported by static analysis
🏁 Script executed:
sed -n '55,125p' src/shared/agent-hook-listener/providers/pi-family-events.ts
sed -n '165,208p' src/shared/program-status/agent-status-projector.ts
sed -n '112,157p' src/renderer/src/components/native-chat/use-native-chat-interactive-send.tsRepository: stablyai/orca
Length of output: 6691
🏁 Script executed:
set -o pipefail
printf '%s\n' '--- OMP status producer references ---'
rg -n -F --glob '*.ts' -e 'tool_execution_start' -e 'resolvePrompt' -e 'projectRoot' -e 'rootStateStartedAt' src/shared src/renderer/src | head -160
printf '%s\n' '--- renderer sender block ---'
nl -ba src/renderer/src/components/native-chat/use-native-chat-interactive-send.ts | sed -n '1,190p'
printf '%s\n' '--- projector block ---'
nl -ba src/shared/program-status/agent-status-projector.ts | sed -n '130,220p'
printf '%s\n' '--- prompt/status listener block ---'
nl -ba src/shared/agent-hook-listener/providers/pi-family-events.ts | sed -n '1,240p'Repository: stablyai/orca
Length of output: 33439
🏁 Script executed:
set -o pipefail
printf '%s\n' '--- OMP listener callers and status timestamp flow ---'
rg -n -F --glob '*.ts' -e 'normalizePiCompatibleEvent(' -e 'lastStatusByPaneKey' -e 'stateStartedAt' src/shared src/renderer/src | head -220
printf '%s\n' '--- listener state and dispatch definitions ---'
rg -n -F --glob '*.ts' -e 'type HookListenerState' -e 'interface HookListenerState' -e 'lastStatusByPaneKey:' -e 'normalizeAgentStatusPayload' src/shared/agent-hook-listener src/shared | head -180
printf '%s\n' '--- prompt field cache ---'
nl -ba src/shared/agent-hook-listener/prompt-fields.ts | sed -n '1,180p'
printf '%s\n' '--- status event applicator ---'
nl -ba src/renderer/src/hooks/ipc-events/agent-status-event-applicator.ts | sed -n '180,270p'Repository: stablyai/orca
Length of output: 42599
🏁 Script executed:
set -o pipefail
printf '%s\n' '--- non-test stateStartedAt assignments ---'
rg -n -F --glob '*.ts' --glob '!*.test.ts' -- 'stateStartedAt' src/shared src/renderer/src
printf '%s\n' '--- hook listener publish callers ---'
rg -n -F --glob '*.ts' --glob '!*.test.ts' -e 'normalizePiCompatibleEvent' -e 'publishAgentStatus' -e 'applyAgentStatus' -e 'stateStartedAt:' src/shared/agent-hook-listener.ts src/shared/agent-hook-status-cache.ts src/shared/agent-status-event* src/shared | head -220
printf '%s\n' '--- main hook listener dispatch ---'
nl -ba src/shared/agent-hook-listener.ts | sed -n '1,340p'
printf '%s\n' '--- hook status cache ---'
nl -ba src/shared/agent-hook-status-cache.ts | sed -n '1,220p'Repository: stablyai/orca
Length of output: 42341
🏁 Script executed:
set -o pipefail
printf '%s\n' '--- lead status fold ---'
nl -ba src/shared/agent-lead-status-fold.ts | sed -n '1,135p'
printf '%s\n' '--- fold callers and status timing construction ---'
rg -n -F --glob '*.ts' --glob '!*.test.ts' -e 'continueAgentLeadStatus' -e 'foldAgentLeadStatus' -e 'agent-lead-status-fold' -e 'stateStartedAt:' -e 'receivedAt:' src/shared src/renderer/src | head -260
printf '%s\n' '--- status mutation and IPC payload path ---'
nl -ba src/shared/agent-status-store-mutation.ts | sed -n '180,310p'
nl -ba src/shared/agent-status-ipc-payload.ts | sed -n '1,125p'
printf '%s\n' '--- tmux hook owner timing path ---'
nl -ba src/shared/tmux-agent-hook-owner.ts | sed -n '95,165p'Repository: stablyai/orca
Length of output: 39933
🏁 Script executed:
set -o pipefail
printf '%s\n' '--- OMP tool fields and event identity ---'
nl -ba src/shared/agent-hook-listener/providers/pi-family-tool-fields.ts | sed -n '1,150p'
rg -n -F --glob '*.ts' -e 'promptInteractionKey' -e 'tool_use_id' -e 'toolUseId' -e 'tool_execution_start' src/shared/agent-hook-listener src/shared/agent-hook-listener.ts src/renderer/src/components/native-chat src/renderer/src/hooks | head -260
printf '%s\n' '--- OMP and Pi-family tests around repeated asks ---'
rg -n -F --glob '*test.ts' -e 'agentType: .omp.' -e 'tool_name: .ask.' -e 'tool_use_id' -e 'promptInteractionKey' -e 'blocked' src/shared/agent-hook-listener* src/shared | head -240
printf '%s\n' '--- status payload identity fields ---'
nl -ba src/shared/agent-hook-listener/listener-event.ts | sed -n '1,180p'
nl -ba src/shared/agent-status-types.ts | sed -n '150,225p'Repository: stablyai/orca
Length of output: 41465
Include a stable OMP ask occurrence in the retry fence.
stateStartedAt is required, so the timestamp-free premise is incorrect. However, the OMP hook publisher preserves stateStartedAt when a new event remains blocked, while updating receivedAt. Two distinct asks with identical question content can therefore produce the same baselineKey.
Do not reset the fence on generic status changes or use receivedAt. That can allow duplicate writes for one ask. Populate the existing promptInteractionKey (or an equivalent stable OMP ask identity) for each ask occurrence and include it in baselineKey. Retransmissions of one ask must retain the same key, while a distinct ask must receive a new key.
There was a problem hiding this comment.
Actionable comments posted: 1
ℹ️ Review info
⚙️ Run configuration
- Configuration used: Repository UI
- Review profile: CHILL
- Plan: Advanced
- Run ID:
0cd29ea8-c850-4a6c-991d-1d31a685231b
📒 Files selected for processing: 9
mobile/src/session/use-mobile-bridge-chat-prompt-writes.tsmobile/src/session/use-mobile-native-chat-answer-send-omp.test.tsmobile/src/session/use-mobile-native-chat-answer-send.tsmobile/src/session/use-mobile-native-chat-controller.tssrc/renderer/src/components/native-chat/use-native-chat-interactive-send.test.tsxsrc/renderer/src/components/native-chat/use-native-chat-interactive-send.tssrc/shared/native-chat-omp-answer-retry-fence.test.tssrc/shared/native-chat-omp-answer-retry-fence.tssrc/shared/native-chat-omp-prompt-identity.ts
Included review availability: This review used your included allowance. Your plan provides up to 10 included reviews per hour; 0 remain after this review.
| const OMP_ANSWER_RETRY_FENCE_CAPACITY = 128 | ||
|
|
||
| export function createOmpAnswerRetryFence(): { | ||
| blocks: (key: string) => boolean | ||
| record: (key: string) => void | ||
| } { | ||
| const attempts = new Set<string>() | ||
| return { | ||
| blocks: (key) => attempts.has(key) || attempts.size >= OMP_ANSWER_RETRY_FENCE_CAPACITY, | ||
| record: (key) => { | ||
| // Retain earlier attempts; reaching capacity must never reopen stale retries. | ||
| if (attempts.size < OMP_ANSWER_RETRY_FENCE_CAPACITY) { | ||
| attempts.add(key) | ||
| } | ||
| } | ||
| } | ||
| } |
There was a problem hiding this comment.
🩺 Stability & Availability | 🟡 Minor | ⚡ Quick win
🔎 Supported by static analysis
🏁 Script executed:
set -u
printf '%s\n' '--- diff summary ---'
git diff --stat 7dfcfbf7356b0a34f12d4c3d2b48d4a2a7dcfdd5 58794aca7f811d2773b178812a9c8c2ed17edb9f -- src/shared/native-chat-omp-answer-retry-fence.ts src/shared/native-chat-omp-answer-retry-fence.test.ts
printf '%s\n' '--- fence source ---'
nl -ba src/shared/native-chat-omp-answer-retry-fence.ts
printf '%s\n' '--- fence test ---'
nl -ba src/shared/native-chat-omp-answer-retry-fence.test.ts
printf '%s\n' '--- all fence references ---'
rg -n -F --glob '!node_modules' --glob '!dist' -e 'createOmpAnswerRetryFence' -e 'OMP_ANSWER_RETRY_FENCE_CAPACITY' -e 'attempts.size' -e 'native-chat-omp-answer-retry-fence' . || test "$?" -eq 1
printf '%s\n' '--- sender lifecycle references ---'
rg -n -F --glob '*.{ts,tsx}' -e 'useState(createOmpAnswerRetryFence)' -e 'fence.blocks' -e 'fence.record' -e 'cleanup' -e 'cancel' renderer/src/components/native-chat mobile/src/session 2>/dev/null || true
printf '%s\n' '--- bounded-memory/capacity language ---'
rg -n -i --glob '*.{ts,tsx,md,json,yml,yaml}' -e 'bounded memory' -e 'memory bound' -e 'capacity' -e 'retry fence' -e 'retry-fence' -e 'max.*attempt' -e 'evict' src renderer mobile docs .github 2>/dev/null || true
printf '%s\n' '--- relevant diff ---'
git diff --unified=80 7dfcfbf7356b0a34f12d4c3d2b48d4a2a7dcfdd5 58794aca7f811d2773b178812a9c8c2ed17edb9f -- src/shared/native-chat-omp-answer-retry-fence.ts src/shared/native-chat-omp-answer-retry-fence.test.tsRepository: stablyai/orca
Length of output: 45657
🏁 Script executed:
printf '%s\n' '--- desktop sender ---'
nl -ba src/renderer/src/components/native-chat/use-native-chat-interactive-send.ts | sed -n '1,125p'
printf '%s\n' '--- mobile sender part 1 ---'
nl -ba mobile/src/session/use-mobile-native-chat-answer-send.ts | sed -n '1,190p'
printf '%s\n' '--- mobile sender part 2 ---'
nl -ba mobile/src/session/use-mobile-native-chat-answer-send.ts | sed -n '190,380p'
printf '%s\n' '--- exact fence key construction and calls ---'
rg -n -C 8 -F --glob '*.{ts,tsx}' -e 'attemptedOmpAsks' -e 'ompAnswer' -e 'askKey' -e 'answerKey' src/renderer/src/components/native-chat mobile/src/session src/shared | head -n 500
printf '%s\n' '--- direct fence contract comments/docs ---'
rg -n -F --glob '*.{ts,tsx,md}' -e 'native-chat-omp-answer-retry-fence' -e 'OMP answer retry fence' -e 'OMP_ANSWER_RETRY_FENCE_CAPACITY' -e 'bounded' -e 'unbounded' src/shared src/renderer/src/components/native-chat mobile/src/session docs 2>/dev/null || trueRepository: stablyai/orca
Length of output: 41676
Remove the capacity guard from both fence operations.
After 128 recorded OMP attempts, blocks rejects every new key. Remove that fail-closed branch and let record retain every key. Keep attempts.has(key) so an uncertain retry remains blocked.
Suggested fix
-const OMP_ANSWER_RETRY_FENCE_CAPACITY = 128
-
export function createOmpAnswerRetryFence(): {
blocks: (key: string) => boolean
record: (key: string) => void
@@
const attempts = new Set<string>()
return {
- blocks: (key) => attempts.has(key) || attempts.size >= OMP_ANSWER_RETRY_FENCE_CAPACITY,
+ blocks: (key) => attempts.has(key),
record: (key) => {
- // Retain earlier attempts; reaching capacity must never reopen stale retries.
- if (attempts.size < OMP_ANSWER_RETRY_FENCE_CAPACITY) {
- attempts.add(key)
- }
+ attempts.add(key)
}
}
}Update the test to expect an unseen key to remain allowed after 200 distinct recorded keys, then verify that recording it makes it blocked.
📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| const OMP_ANSWER_RETRY_FENCE_CAPACITY = 128 | |
| export function createOmpAnswerRetryFence(): { | |
| blocks: (key: string) => boolean | |
| record: (key: string) => void | |
| } { | |
| const attempts = new Set<string>() | |
| return { | |
| blocks: (key) => attempts.has(key) || attempts.size >= OMP_ANSWER_RETRY_FENCE_CAPACITY, | |
| record: (key) => { | |
| // Retain earlier attempts; reaching capacity must never reopen stale retries. | |
| if (attempts.size < OMP_ANSWER_RETRY_FENCE_CAPACITY) { | |
| attempts.add(key) | |
| } | |
| } | |
| } | |
| } | |
| export function createOmpAnswerRetryFence(): { | |
| blocks: (key: string) => boolean | |
| record: (key: string) => void | |
| } { | |
| const attempts = new Set<string>() | |
| return { | |
| blocks: (key) => attempts.has(key), | |
| record: (key) => { | |
| attempts.add(key) | |
| } | |
| } | |
| } |
Use a visible terminal action for OMP ask dialogs and refuse automated card answer/cancel writes. Preserve structured-session responses and OMP multi-select parsing. Co-authored-by: jairuspace <jairusbpace@gmail.com> Co-authored-by: Daniel Lopez <dafeloru@gmail.com>
There was a problem hiding this comment.
Actionable comments posted: 1
ℹ️ Review info
⚙️ Run configuration
- Configuration used: Repository UI
- Review profile: CHILL
- Plan: Advanced
- Run ID:
719410ea-4a86-44f6-b9e9-b45552a6c389
📒 Files selected for processing: 22
mobile/src/session/MobileNativeChatOverlay.test.tsmobile/src/session/MobileNativeChatOverlay.tsxmobile/src/session/MobileNativeChatPromptCard.test.tsxmobile/src/session/MobileNativeChatPromptCard.tsxmobile/src/session/MobileNativeChatView.prompt-presentation.test.tsmobile/src/session/MobileNativeChatView.tsxmobile/src/session/MobileSessionActiveContent.tsxmobile/src/session/use-mobile-bridge-chat-prompt-writes.tsmobile/src/session/use-mobile-native-chat-answer-send-omp.test.tsmobile/src/session/use-mobile-native-chat-answer-send.tsmobile/src/session/use-mobile-native-chat-cancel-ask.test.tsmobile/src/session/use-mobile-native-chat-cancel-ask.tssrc/renderer/src/components/native-chat/NativeChatInteractiveCard.test.tsxsrc/renderer/src/components/native-chat/NativeChatInteractiveCard.tsxsrc/renderer/src/components/native-chat/NativeChatResolvedView.reveal-latest.test.tsxsrc/renderer/src/components/native-chat/NativeChatResolvedView.tsxsrc/renderer/src/components/native-chat/use-native-chat-interactive-send.test.tsxsrc/renderer/src/components/native-chat/use-native-chat-interactive-send.tssrc/renderer/src/i18n/locales/en.jsonsrc/shared/native-chat-agent-support.tssrc/shared/native-chat-ask.test.tssrc/shared/native-chat-ask.ts
💤 Files with no reviewable changes: 1
- src/shared/native-chat-ask.ts
Included review availability: This review used your included allowance. Your plan provides up to 10 included reviews per hour; 3 remain after this review.
| let tree: ReactTestRenderer | null = null | ||
| await act(async () => { | ||
| tree = create( | ||
| createElement(MobileNativeChatPromptCard, { | ||
| ask, | ||
| requiresTerminal: true, | ||
| onAnswerAsk, | ||
| onCancelAsk, | ||
| onDismissAsk, | ||
| onAnswerInTerminal: available ? onAnswerInTerminal : undefined | ||
| }) | ||
| ) | ||
| }) | ||
| const buttons = tree!.root.findAllByType('Pressable') | ||
| expect(buttons).toHaveLength(1) | ||
| expect(buttons[0]!.props.accessibilityLabel).toBe('Answer in terminal') | ||
| expect(buttons[0]!.props.disabled).toBe(!available) | ||
| expect(tree!.root.findAllByType('TextInput')).toHaveLength(0) | ||
| if (available) { | ||
| act(() => buttons[0]!.props.onPress()) | ||
| } | ||
| expect(onAnswerInTerminal).toHaveBeenCalledTimes(available ? 1 : 0) | ||
| expect(onAnswerAsk).not.toHaveBeenCalled() | ||
| expect(onCancelAsk).not.toHaveBeenCalled() | ||
| expect(onDismissAsk).not.toHaveBeenCalled() | ||
| act(() => tree!.unmount()) |
There was a problem hiding this comment.
🎯 Functional Correctness | 🟠 Major | ⚡ Quick win
Fix the narrowed tree variable. It breaks typecheck:tests and fails the "Mobile verify" job.
Line 37 declares let tree: ReactTestRenderer | null = null. The code assigns tree only inside the act callback. TypeScript narrows tree to null after the callback. At Lines 50 and 54, tree! therefore has type never, and .root is a type error. This matches the pipeline failure "Test file no longer typechecks".
Store the renderer on an object property. Property reads are not narrowed by the initializer. The repo already uses this pattern in agent-history-current-host-capability.test.tsx (rendered.tree = create(...)).
💚 Proposed fix
--- "a/mobile/src/session/MobileNativeChatPromptCard.test.tsx"
+++ "b/mobile/src/session/MobileNativeChatPromptCard.test.tsx"
@@ -34,32 +34,32 @@
const onAnswerAsk = vi.fn(async () => true)
const onCancelAsk = vi.fn(async () => true)
const onDismissAsk = vi.fn()
- let tree: ReactTestRenderer | null = null
+ const rendered: { tree: ReactTestRenderer | null } = { tree: null }
await act(async () => {
- tree = create(
+ rendered.tree = create(
createElement(MobileNativeChatPromptCard, {
ask,
requiresTerminal: true,
onAnswerAsk,
onCancelAsk,
onDismissAsk,
onAnswerInTerminal: available ? onAnswerInTerminal : undefined
})
)
})
- const buttons = tree!.root.findAllByType('Pressable')
+ const buttons = rendered.tree!.root.findAllByType('Pressable')
expect(buttons).toHaveLength(1)
expect(buttons[0]!.props.accessibilityLabel).toBe('Answer in terminal')
expect(buttons[0]!.props.disabled).toBe(!available)
- expect(tree!.root.findAllByType('TextInput')).toHaveLength(0)
+ expect(rendered.tree!.root.findAllByType('TextInput')).toHaveLength(0)
if (available) {
act(() => buttons[0]!.props.onPress())
}
expect(onAnswerInTerminal).toHaveBeenCalledTimes(available ? 1 : 0)
expect(onAnswerAsk).not.toHaveBeenCalled()
expect(onCancelAsk).not.toHaveBeenCalled()
expect(onDismissAsk).not.toHaveBeenCalled()
- act(() => tree!.unmount())
+ act(() => rendered.tree!.unmount())
}
)
📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| let tree: ReactTestRenderer | null = null | |
| await act(async () => { | |
| tree = create( | |
| createElement(MobileNativeChatPromptCard, { | |
| ask, | |
| requiresTerminal: true, | |
| onAnswerAsk, | |
| onCancelAsk, | |
| onDismissAsk, | |
| onAnswerInTerminal: available ? onAnswerInTerminal : undefined | |
| }) | |
| ) | |
| }) | |
| const buttons = tree!.root.findAllByType('Pressable') | |
| expect(buttons).toHaveLength(1) | |
| expect(buttons[0]!.props.accessibilityLabel).toBe('Answer in terminal') | |
| expect(buttons[0]!.props.disabled).toBe(!available) | |
| expect(tree!.root.findAllByType('TextInput')).toHaveLength(0) | |
| if (available) { | |
| act(() => buttons[0]!.props.onPress()) | |
| } | |
| expect(onAnswerInTerminal).toHaveBeenCalledTimes(available ? 1 : 0) | |
| expect(onAnswerAsk).not.toHaveBeenCalled() | |
| expect(onCancelAsk).not.toHaveBeenCalled() | |
| expect(onDismissAsk).not.toHaveBeenCalled() | |
| act(() => tree!.unmount()) | |
| const rendered: { tree: ReactTestRenderer | null } = { tree: null } | |
| await act(async () => { | |
| rendered.tree = create( | |
| createElement(MobileNativeChatPromptCard, { | |
| ask, | |
| requiresTerminal: true, | |
| onAnswerAsk, | |
| onCancelAsk, | |
| onDismissAsk, | |
| onAnswerInTerminal: available ? onAnswerInTerminal : undefined | |
| }) | |
| ) | |
| }) | |
| const buttons = rendered.tree!.root.findAllByType('Pressable') | |
| expect(buttons).toHaveLength(1) | |
| expect(buttons[0]!.props.accessibilityLabel).toBe('Answer in terminal') | |
| expect(buttons[0]!.props.disabled).toBe(!available) | |
| expect(rendered.tree!.root.findAllByType('TextInput')).toHaveLength(0) | |
| if (available) { | |
| act(() => buttons[0]!.props.onPress()) | |
| } | |
| expect(onAnswerInTerminal).toHaveBeenCalledTimes(available ? 1 : 0) | |
| expect(onAnswerAsk).not.toHaveBeenCalled() | |
| expect(onCancelAsk).not.toHaveBeenCalled() | |
| expect(onDismissAsk).not.toHaveBeenCalled() | |
| act(() => rendered.tree!.unmount()) |
Source: Pipeline failures
OMP question cards could silently submit the highlighted option, toggle the wrong checkbox state, or answer the wrong question after terminal navigation. Desktop and mobile now show Answer in terminal for terminal-backed OMP questions. Opening the terminal preserves the active dialog and draft; card answer/cancel callbacks refuse to write terminal input.
Automatic OMP card answering is deferred. Refs #26692 and #26198; this PR does not close the remaining request for semantic card answering.
Changes
multiflag in canonical status and transcript parsing.Why terminal input cannot safely submit these answers
Verification against the actual selector implementation reproduces three independent state ambiguities:
The current public local extension API can create a dialog but cannot resolve an existing ask by ID. Semantic RPC and collaboration responses belong to different process/session transports. Reusing them for ordinary terminal sessions would require a separate integration.
Verification
d922b3f3240found no blocking findings and independently passed the same 99 mobile and 76 desktop/shared tests.7bbfc88e0e4only fixes typed component selectors in two test fixtures; no production delta from the independent review. Both affected suites (10 tests) and the native mobile test-typecheck ratchet pass.Community credit
Follow-up to @jairuspace's #26688 and @daniellp-rubio's #29184. Both contributors remain credited through
Co-authored-bycommit trailers. The final scope is the safe terminal fallback rather than automatic keyboard submission.Risk
LOW for this narrowed scope, independently approved at
d922b3f3240; full final-head PR CI and aggregate pass. Optional WSL distro CI remains queued. The intentional limitation is that terminal-backed OMP questions must be answered in terminal view. Automatic card answering still needs an active-dialog-ID-bound semantic answer API.