Conversation
…tions, and SwiftUI P0 (Critical): - Skip selected tab during memory eviction to prevent visible refresh on window re-focus - Wrap reconnectDriver in trackOperation to prevent health monitor race condition P1 (High): - Cancel redisDatabaseSwitchTask in coordinator teardown - Add deinit to AIChatViewModel to cancel streaming task - Replace 5x fire-and-forget SSH tunnel closes with error logging - Log schema/database switch errors instead of silently discarding - Re-fetch session for driver disconnect to avoid stale reference - Add WindowAccessor NSViewRepresentable for reliable window capture - Use viewWindow instead of NSApp.keyWindow in command actions P2 (Medium): - Fix weak self in LicenseManager/AnalyticsService periodic task loops - Remove imprecise global changeManager check from eviction scheduling - Add query staleness detection so stuck queries don't mask dead connections - Log warning for synchronous plugin loading fallback - Add deinit to AppSettingsManager for observer cleanup - Split clearChanges/clearChangesAndUndoHistory for undo preservation - Make TabDiskActor.save() throw so callers can handle disk errors - Remove fragile subtitle-based window matching fallback - Use viewDidMoveToWindow for synchronous window capture P3 (Low): - Remove stale isKeyWindow snapshot from configureWindow - Centralize window identifier constants with exact matching - Remove unnecessary DispatchQueue.main.async in AppDelegate handlers - Add warning log for WindowLifecycleMonitor re-registration - Unify exponential backoff via shared ExponentialBackoff utility - Return startup command failures from executeStartupCommands - Replace recursive asyncAfter with cancellable Task loop
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (2)
🚧 Files skipped from review as they are similar to previous changes (1)
📝 WalkthroughWalkthroughFixes and refinements across connection, window, persistence, and change-tracking code: MongoDB SRV authSource and TLS handling, improved stuck-query detection and SSH tunnel/logging, preserved undo history across tab switches, exact window-identifier matching, throwing tab-disk saves, centralized backoff, and various lifecycle/cleanup fixes. Changes
Sequence Diagram(s)sequenceDiagram
participant App as App / UI
participant DBM as DatabaseManager
participant Driver as DatabaseDriver / PluginManager
participant SSH as SSH Tunnel
participant HB as ConnectionHealthMonitor
rect rgba(200,200,255,0.5)
App->>DBM: request reconnect / execute query
DBM->>DBM: trackOperation(sessionId) (set queryStartTimes)
DBM->>Driver: createDriver(for:)
Driver-->>DBM: driver OR loadPendingPlugins() (sync)
DBM->>SSH: open tunnel (if needed)
SSH-->>DBM: success / throw (on error -> log warning)
DBM->>Driver: run query
Driver-->>DBM: query in-flight / result
DBM->>HB: health pings (allow if queryStart exceeded timeout)
HB-->>DBM: ping result / mark stuck
end
Estimated code review effort🎯 4 (Complex) | ⏱️ ~60 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 3✅ Passed checks (3 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 4
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (2)
TablePro/Views/Main/MainContentCommandActions.swift (1)
377-382:⚠️ Potential issue | 🟠 MajorClose the owning window, not whichever window is key later.
discardAndClose()still falls through toperformClose(), and that helper resolvesNSApp.keyWindowinstead of this instance'swindow. If focus changes while the async save/discard prompt is up, the close can hit a different native window tab than the one that initiated the command.🔧 Suggested direction
- private func performClose() { - guard let keyWindow = NSApp.keyWindow else { return } - let tabbedWindows = keyWindow.tabbedWindows ?? [keyWindow] + private func performClose() { + guard let window else { return } + let tabbedWindows = window.tabbedWindows ?? [window] if tabbedWindows.count > 1 { - keyWindow.close() + window.close() } else if coordinator?.tabManager.tabs.isEmpty == true { - keyWindow.close() + window.close() } else {🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@TablePro/Views/Main/MainContentCommandActions.swift` around lines 377 - 382, discardAndClose() currently calls performClose(), which uses NSApp.keyWindow so if focus changes during the async prompt we may close the wrong window; change the logic so discardAndClose() closes this view's owning window directly (use self.window?.close() or pass the specific NSWindow instance into performClose) instead of relying on NSApp.keyWindow; update performClose signature if needed (e.g., performClose(window: NSWindow?)) and replace any NSApp.keyWindow lookups with the explicit window parameter or use self.window within discardAndClose to ensure the originating window/tab is closed.TablePro/Views/Main/Extensions/MainContentCoordinator+SaveChanges.swift (1)
231-235:⚠️ Potential issue | 🟠 MajorBind post-save cleanup to the tab that started the save.
This runs after async I/O, but it clears
changeManager, resetspendingChanges, and re-runs the query using the tab that happens to be selected at completion time. If the user switches tabs while the commit is running, the new tab gets marked clean and loses its undo history even though it was never saved.🔧 Suggested scoping
+ let savedTabId = tabManager.selectedTabId Task { `@MainActor` in let overallStartTime = Date() do { guard let driver = DatabaseManager.shared.driver(for: connectionId) else { @@ - changeManager.clearChangesAndUndoHistory() - if let index = tabManager.selectedTabIndex { + if tabManager.selectedTabId == savedTabId { + changeManager.clearChangesAndUndoHistory() + } + if let savedTabId, + let index = tabManager.tabs.firstIndex(where: { $0.id == savedTabId }) { tabManager.tabs[index].pendingChanges = TabPendingChanges() tabManager.tabs[index].errorMessage = nil } @@ - if tabManager.selectedTabIndex != nil && !tabManager.tabs.isEmpty { + if tabManager.selectedTabId == savedTabId && !tabManager.tabs.isEmpty { runQuery() }Also applies to: 262-264
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@TablePro/Views/Main/Extensions/MainContentCoordinator`+SaveChanges.swift around lines 231 - 235, The post-save cleanup is currently using tabManager.selectedTabIndex which can change during async I/O; instead capture the originating tab (e.g., store the starting selectedTabIndex or the tab's unique identifier before awaiting) and perform changeManager.clearChangesAndUndoHistory() and the resets on tabManager.tabs[originatingIndex] (or locate by captured tab id) only if that tab still exists, leaving other tabs untouched; apply the same change to the other cleanup site (the block around the symbols at lines referenced, e.g., the second occurrence that currently sets tabManager.tabs[index].pendingChanges and .errorMessage) so undo history and pendingChanges are tied to the tab that initiated the save.
🧹 Nitpick comments (4)
TablePro/Views/Components/WindowAccessor.swift (2)
7-17: Add explicit access control modifiers.Per coding guidelines, types should have explicit access control. The
struct WindowAccessorandfinal class WindowAccessorViewshould specify their intended visibility.♻️ Proposed fix
-struct WindowAccessor: NSViewRepresentable { +internal struct WindowAccessor: NSViewRepresentable { var onWindow: (NSWindow) -> Void func makeNSView(context: Context) -> WindowAccessorView {-final class WindowAccessorView: NSView { +internal final class WindowAccessorView: NSView { var onWindow: ((NSWindow) -> Void)?As per coding guidelines: "Always specify access control explicitly (private, internal, public) on extensions and types."
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@TablePro/Views/Components/WindowAccessor.swift` around lines 7 - 17, Add explicit access control to the types and their API: declare the struct WindowAccessor with an explicit access level (e.g., internal or public as appropriate), and likewise mark final class WindowAccessorView with the chosen access level; also apply explicit access control to the onWindow property and the makeNSView/updateNSView methods (and any other members) so the intended visibility is clear and consistent.
22-27: Consider guarding against multiple invocations.
viewDidMoveToWindowcan be called multiple times if the view hierarchy changes (e.g., view moves between windows). The current implementation will invokeonWindoweach time. If the callback should only fire once, consider adding a guard.♻️ Optional: fire callback only once
internal final class WindowAccessorView: NSView { var onWindow: ((NSWindow) -> Void)? + private var didCapture = false override func viewDidMoveToWindow() { super.viewDidMoveToWindow() - if let window { + if let window, !didCapture { + didCapture = true onWindow?(window) } } }This is optional — if the callback is idempotent or intentionally supports re-registration, the current behavior is fine.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@TablePro/Views/Components/WindowAccessor.swift` around lines 22 - 27, viewDidMoveToWindow can be called multiple times and currently always invokes onWindow; add a guard boolean (e.g., private var hasInvokedOnWindow = false) and in viewDidMoveToWindow only call onWindow when window != nil && !hasInvokedOnWindow, then set hasInvokedOnWindow = true; if you want to allow re-firing when the view leaves and re-enters a window, clear hasInvokedOnWindow when window == nil instead of permanently setting it.TablePro/AppDelegate+WindowConfig.swift (1)
319-324: Inconsistent window matching incloseRestoredMainWindows().Line 321 still uses
.contains("main")for matching, which is inconsistent with the new exact matching approach elsewhere in this file. Consider usingWindowId.mainfor consistency:Suggested fix
func closeRestoredMainWindows() { DispatchQueue.main.async { - for window in NSApp.windows where window.identifier?.rawValue.contains("main") == true { + for window in NSApp.windows where window.identifier?.rawValue == WindowId.main { window.close() } } }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@TablePro/AppDelegate`+WindowConfig.swift around lines 319 - 324, The window matching in closeRestoredMainWindows() is inconsistent: replace the substring check using .contains("main") with an exact match to WindowId.main (e.g., compare window.identifier?.rawValue == WindowId.main.rawValue or == WindowId.main) so it uses the same precise identifier logic as elsewhere; update the loop in closeRestoredMainWindows() that iterates NSApp.windows to perform that exact equality check and close matching windows.TablePro/Core/Utilities/Connection/ExponentialBackoff.swift (1)
4-17: Add explicit access control.Per coding guidelines, access control should be specified explicitly on types. The enum defaults to
internal, but should be explicitly marked.Suggested fix
/// Shared exponential backoff calculator for reconnection delays. -enum ExponentialBackoff { +internal enum ExponentialBackoff { private static let seeds: [TimeInterval] = [2, 4, 8]As per coding guidelines: "Always specify access control explicitly (private, internal, public) on extensions and types."
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@TablePro/Core/Utilities/Connection/ExponentialBackoff.swift` around lines 4 - 17, The enum ExponentialBackoff currently has implicit access control; explicitly declare its access level by changing the type declaration to include access control (e.g., "internal enum ExponentialBackoff") so the type follows the project's guideline; keep the existing private static let seeds and the static func delay as-is (they retain their explicit/private modifiers), ensuring only the enum declaration is updated.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@CHANGELOG.md`:
- Around line 10-13: The Unreleased section currently contains MongoDB bullets
that belong to a different PR; update CHANGELOG.md under the [Unreleased]
(Added/Fixed/Changed) section by either removing the two MongoDB bullets
("MongoDB Atlas connections failing to authenticate" and "MongoDB TLS
certificate verification skipped for SRV connections") or replacing them with
this PR's actual notable change(s) (e.g., "window/memory/connection anti-pattern
cleanup" or a succinct summary of the refactor that fixed those anti-patterns);
ensure the entry clearly names the change and the PR number or author per the
repo's changelog guidelines.
In `@Plugins/MongoDBDriverPlugin/MongoDBConnection.swift`:
- Around line 208-213: The change restricts setting
tlsAllowInvalidCertificates=true to sslMode "Preferred" (using sslMode,
sslEnabled, params in MongoDBConnection.swift), which is a security improvement;
add a concise inline comment above the sslEnabled block explaining that
tlsAllowInvalidCertificates is intentionally only set for "Preferred" to avoid
bypassing certificate validation in "Required" (and that users with self-signed
certs should temporarily use "Preferred" or install proper CA certs and migrate
to "Verify CA"/"Verify Identity"), and update any user-facing docs or README to
mention this behavioral change and migration guidance for self-signed
certificate setups.
In `@TablePro/Views/Main/Extensions/MainContentCoordinator`+MultiStatement.swift:
- Around line 137-138: The code is clearing shared change state unconditionally
after a background multi-statement completion; wrap the mutation of
changeManager (calls to clearChangesAndUndoHistory() and incrementing
reloadVersion) in a guard that checks the originating tabId resolved by the task
matches the currently selected/current tab tracked by the
coordinator/changeManager (e.g., compare originatingTabId to
changeManager.currentTab?.id or coordinator.selectedTabId) and only perform the
clear/reload when they match so switching tabs mid-run won't wipe another tab's
undo history.
In `@TablePro/Views/Main/Extensions/MainContentCoordinator`+QueryHelpers.swift:
- Around line 195-197: The code unconditionally calls
changeManager.clearChangesAndUndoHistory() inside applyPhase1Result, which can
clear another tab's undo/redo when a background query finishes; guard this call
so it only runs when the phase1 result applies to the currently visible/selected
tab (e.g., compare the result's target tab identifier or an isSelected flag for
the tab being updated) and only then, and still check isEditable &&
!changeManager.hasChanges before clearing; update applyPhase1Result to perform
that selected-tab check before invoking
changeManager.clearChangesAndUndoHistory().
---
Outside diff comments:
In `@TablePro/Views/Main/Extensions/MainContentCoordinator`+SaveChanges.swift:
- Around line 231-235: The post-save cleanup is currently using
tabManager.selectedTabIndex which can change during async I/O; instead capture
the originating tab (e.g., store the starting selectedTabIndex or the tab's
unique identifier before awaiting) and perform
changeManager.clearChangesAndUndoHistory() and the resets on
tabManager.tabs[originatingIndex] (or locate by captured tab id) only if that
tab still exists, leaving other tabs untouched; apply the same change to the
other cleanup site (the block around the symbols at lines referenced, e.g., the
second occurrence that currently sets tabManager.tabs[index].pendingChanges and
.errorMessage) so undo history and pendingChanges are tied to the tab that
initiated the save.
In `@TablePro/Views/Main/MainContentCommandActions.swift`:
- Around line 377-382: discardAndClose() currently calls performClose(), which
uses NSApp.keyWindow so if focus changes during the async prompt we may close
the wrong window; change the logic so discardAndClose() closes this view's
owning window directly (use self.window?.close() or pass the specific NSWindow
instance into performClose) instead of relying on NSApp.keyWindow; update
performClose signature if needed (e.g., performClose(window: NSWindow?)) and
replace any NSApp.keyWindow lookups with the explicit window parameter or use
self.window within discardAndClose to ensure the originating window/tab is
closed.
---
Nitpick comments:
In `@TablePro/AppDelegate`+WindowConfig.swift:
- Around line 319-324: The window matching in closeRestoredMainWindows() is
inconsistent: replace the substring check using .contains("main") with an exact
match to WindowId.main (e.g., compare window.identifier?.rawValue ==
WindowId.main.rawValue or == WindowId.main) so it uses the same precise
identifier logic as elsewhere; update the loop in closeRestoredMainWindows()
that iterates NSApp.windows to perform that exact equality check and close
matching windows.
In `@TablePro/Core/Utilities/Connection/ExponentialBackoff.swift`:
- Around line 4-17: The enum ExponentialBackoff currently has implicit access
control; explicitly declare its access level by changing the type declaration to
include access control (e.g., "internal enum ExponentialBackoff") so the type
follows the project's guideline; keep the existing private static let seeds and
the static func delay as-is (they retain their explicit/private modifiers),
ensuring only the enum declaration is updated.
In `@TablePro/Views/Components/WindowAccessor.swift`:
- Around line 7-17: Add explicit access control to the types and their API:
declare the struct WindowAccessor with an explicit access level (e.g., internal
or public as appropriate), and likewise mark final class WindowAccessorView with
the chosen access level; also apply explicit access control to the onWindow
property and the makeNSView/updateNSView methods (and any other members) so the
intended visibility is clear and consistent.
- Around line 22-27: viewDidMoveToWindow can be called multiple times and
currently always invokes onWindow; add a guard boolean (e.g., private var
hasInvokedOnWindow = false) and in viewDidMoveToWindow only call onWindow when
window != nil && !hasInvokedOnWindow, then set hasInvokedOnWindow = true; if you
want to allow re-firing when the view leaves and re-enters a window, clear
hasInvokedOnWindow when window == nil instead of permanently setting it.
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 95412b10-51bd-4838-867f-93f159d74c82
📒 Files selected for processing (27)
CHANGELOG.mdPlugins/MongoDBDriverPlugin/MongoDBConnection.swiftTablePro/AppDelegate+WindowConfig.swiftTablePro/Core/ChangeTracking/DataChangeManager.swiftTablePro/Core/Database/ConnectionHealthMonitor.swiftTablePro/Core/Database/DatabaseDriver.swiftTablePro/Core/Database/DatabaseManager.swiftTablePro/Core/Services/Infrastructure/AnalyticsService.swiftTablePro/Core/Services/Infrastructure/TabPersistenceCoordinator.swiftTablePro/Core/Services/Infrastructure/WindowLifecycleMonitor.swiftTablePro/Core/Services/Licensing/LicenseManager.swiftTablePro/Core/Storage/AppSettingsManager.swiftTablePro/Core/Storage/TabDiskActor.swiftTablePro/Core/Utilities/Connection/ExponentialBackoff.swiftTablePro/ViewModels/AIChatViewModel.swiftTablePro/Views/Components/WindowAccessor.swiftTablePro/Views/Connection/WelcomeWindowView.swiftTablePro/Views/Main/Extensions/MainContentCoordinator+ChangeGuard.swiftTablePro/Views/Main/Extensions/MainContentCoordinator+Discard.swiftTablePro/Views/Main/Extensions/MainContentCoordinator+MultiStatement.swiftTablePro/Views/Main/Extensions/MainContentCoordinator+QueryHelpers.swiftTablePro/Views/Main/Extensions/MainContentCoordinator+Refresh.swiftTablePro/Views/Main/Extensions/MainContentCoordinator+SaveChanges.swiftTablePro/Views/Main/MainContentCommandActions.swiftTablePro/Views/Main/MainContentCoordinator.swiftTablePro/Views/Main/MainContentView.swiftdocs/development/anti-patterns-tracker.md
TablePro/Views/Main/Extensions/MainContentCoordinator+MultiStatement.swift
Outdated
Show resolved
Hide resolved
TablePro/Views/Main/Extensions/MainContentCoordinator+QueryHelpers.swift
Outdated
Show resolved
Hide resolved
- WindowAccessor: deduplicate viewDidMoveToWindow calls, update closure in updateNSView - AIChatViewModel: add comment explaining nonisolated(unsafe) requirement - CHANGELOG: add user-visible behavioral changes under [Unreleased]
Prevent clearing another tab's undo state when a background query or multi-statement execution completes after the user has switched tabs.
datlechin
added a commit
that referenced
this pull request
Mar 24, 2026
… match SwiftUI appends -AppWindow-N to the WindowGroup id, so window.identifier is "main-AppWindow-1" not "main". The exact match from PR #441 broke isMainWindow, preventing tab grouping and window lifecycle handling.
datlechin
added a commit
that referenced
this pull request
Mar 24, 2026
* fix: reduce memory retention after closing tabs - Clear changeManager state and pluginDriver reference in teardown - Cancel redisDatabaseSwitchTask in teardown - Clear cachedTableColumnTypes/Names, tableMetadata, filterState in teardown - Release editor closures and heavy state (tree-sitter, highlighter) on destroy - Add releaseHeavyState() to TextViewController for early resource cleanup - Make InMemoryRowProvider.rowBuffer weak with safe fallback - Add releaseData() to InMemoryRowProvider for explicit cleanup - Clear tabProviderCache, sortCache, cachedChangeManager in onTeardown - Hint malloc to return freed pages after disconnect - Add deinit logging for RowBuffer and QueryTabManager * fix: set isKeyWindow when WindowAccessor captures window WindowAccessor.viewDidMoveToWindow fires after didBecomeKeyNotification, so isKeyWindow was never set to true. This blocked sidebar table clicks (guarded by isKeyWindow) and caused new tabs to open as separate windows instead of attaching to the connection's tab group. * fix: explicitly attach new windows to existing tab group openWindow creates the window before tabbingIdentifier is set, so macOS automatic tabbing doesn't group them. Use addTabbedWindow to explicitly attach new windows to the connection's existing tab group. * fix: add diagnostic logging for tab grouping * fix: add NSLog diagnostic for windowDidBecomeKey * fix: match SwiftUI window identifiers with hasPrefix instead of exact match SwiftUI appends -AppWindow-N to the WindowGroup id, so window.identifier is "main-AppWindow-1" not "main". The exact match from PR #441 broke isMainWindow, preventing tab grouping and window lifecycle handling. * fix: use prefix matching for window identifiers across all call sites SwiftUI appends -AppWindow-N to WindowGroup IDs. Fixed remaining spots that used .contains("main") or exact match: - closeRestoredMainWindows: use isMainWindow helper - closeWindows(withId:): prefix match instead of contains - ContentView notification filter: prefix match instead of contains * fix: address code review feedback - Use queue: .main instead of assumeIsolated in DataGridView teardown observer - Apply hasPrefix matching to isWelcomeWindow and isConnectionFormWindow - Gate RowBuffer/QueryTabManager deinit logging behind #if DEBUG - Add CHANGELOG entries for memory, tab grouping, and sidebar fixes * fix: guard buffer writes against nil to avoid mutating shared emptyBuffer * docs: simplify changelog entries
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
P0 - Critical (2 fixes)
reconnectDriverintrackOperationso health monitor pings don't race on non-thread-safe driversP1 - High (7 fixes)
redisDatabaseSwitchTaskin coordinator teardown (task leak)deinittoAIChatViewModelto cancel streaming taskdo/catch+ warning logstry?silent discardWindowAccessorNSViewRepresentable for reliable window capture (replaces fragileNSApp.keyWindow)viewWindowfrom WindowAccessor in command actions setupP2 - Medium (9 fixes)
LicenseManager/AnalyticsServiceperiodic loopschangeManager.hasChangespre-check from eviction (per-tab check is sufficient)deinittoAppSettingsManagerfor accessibility observer cleanupclearChanges()/clearChangesAndUndoHistory()for undo stack preservationTabDiskActor.save()throw so callers can handle disk errorsWindowLifecycleMonitorviewDidMoveToWindowfor synchronous window capture (eliminates notification timing race)P3 - Low (8 fixes)
isKeyWindowsnapshot; notification handlers are sole source of truth.contains())DispatchQueue.main.asyncin AppDelegate notification handlersWindowLifecycleMonitorre-registration with different windowExponentialBackoffutility (unifies DatabaseManager + HealthMonitor)executeStartupCommandsnow returns failures via@discardableResultasyncAfterpolling with cancellableTaskloopTest plan
viewWindowreferenceSummary by CodeRabbit
New Features
Bug Fixes
Reliability