You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Audit follow-up: remove IPC bridge surface with zero consumers in the renderer/overlay/e2e. All three items are safe deletions verified against current main (HEAD d20c9fc).
Findings
1. window:openSettings push has no sender since #390
apps/desktop/src/preload/preload.ts:1017-1018 still exposes appWindow.subscribeOpenSettings; apps/desktop/src/renderer/app-shell-effects.ts:326 subscribes (unsubscribed at :336).
⌘, is already handled by the renderer keydown handler (app-shell-effects.ts:279-281), so no functional loss.
2. dailyReview.list/get/delete duplicate the live *Archive variants
preload/preload.ts:968-984: list/listArchives → same daily-review:list channel; get/getArchive and delete/deleteArchive likewise.
Renderer only calls listArchives/getArchive (apps/desktop/src/renderer/app-shell-daily-review-bridge.ts:18,23). delete has no consumer in either spelling — the whole daily-review:delete channel is renderer-dead (archive-store retention cleanup calls store.deleteArchive directly).
3. cursor-subscription:logout is not gated by the experimental flag
apps/desktop/src/main/subscription-ipc-main.ts:408-411 registers cursor-subscription:logout unconditionally, unlike the other 7 cursor-subscription:* channels gated by MAKA_CURSOR_SUBSCRIPTION_EXPERIMENTAL !== '0' (oauth/cursor-subscription-helpers.ts:79-80).
Proposed fix
Delete subscribeOpenSettings from preload + the subscribeOpenSettings/handleOpenSettings wiring in app-shell-effects.ts (keep the keydown handler).
Delete dailyReview.list/get/delete aliases from preload (and bridge-contract.d.ts if listed); optionally remove the daily-review:delete channel end-to-end.
Gate cursor-subscription:logout with the same experimental flag, or remove it with the rest of the cursor surface (see the decision issue for the cursor subscription surface).
Summary
Audit follow-up: remove IPC bridge surface with zero consumers in the renderer/overlay/e2e. All three items are safe deletions verified against current
main(HEAD d20c9fc).Findings
1.
window:openSettingspush has no sender since #390apps/desktop/src/preload/preload.ts:1017-1018still exposesappWindow.subscribeOpenSettings;apps/desktop/src/renderer/app-shell-effects.ts:326subscribes (unsubscribed at :336).设置…/CommandOrControl+,inmain-window.ts) was deleted in chore(ui): track Storybook-driven UI polish after file-boundary governance #390 (git show 5d696e463). No sender exists anywhere inmain/or the builtdist/main.app-shell-effects.ts:279-281), so no functional loss.2.
dailyReview.list/get/deleteduplicate the live*Archivevariantspreload/preload.ts:968-984:list/listArchives→ samedaily-review:listchannel;get/getArchiveanddelete/deleteArchivelikewise.listArchives/getArchive(apps/desktop/src/renderer/app-shell-daily-review-bridge.ts:18,23).deletehas no consumer in either spelling — the wholedaily-review:deletechannel is renderer-dead (archive-store retention cleanup callsstore.deleteArchivedirectly).3.
cursor-subscription:logoutis not gated by the experimental flagapps/desktop/src/main/subscription-ipc-main.ts:408-411registerscursor-subscription:logoutunconditionally, unlike the other 7cursor-subscription:*channels gated byMAKA_CURSOR_SUBSCRIPTION_EXPERIMENTAL !== '0'(oauth/cursor-subscription-helpers.ts:79-80).Proposed fix
subscribeOpenSettingsfrom preload + thesubscribeOpenSettings/handleOpenSettingswiring inapp-shell-effects.ts(keep the keydown handler).dailyReview.list/get/deletealiases from preload (andbridge-contract.d.tsif listed); optionally remove thedaily-review:deletechannel end-to-end.cursor-subscription:logoutwith the same experimental flag, or remove it with the rest of the cursor surface (see the decision issue for the cursor subscription surface).Owner
main process / preload (IPC boundary).
Acceptance