https://developer.apple.com/documentation/NetworkExtension/filtering-network-traffic App example not auto reconnect after network extension crush. what need to add for auto reconnect when network extension restart?

Hi, happy new year, I'm a Product Manager for a communications app that's currently in testflight. We requested the com.apple.developer.usernotifications.filtering entitlement on December 3rd, and have yet to receive a response from Apple. I understand that the holiday break may have gotten in the way, however it feels like we were lost in the queue as it's been 6 weeks with no response. Our app owner has checked-in inside appstoreconnect but has not received anything back. Is this common? Is there any process for getting a status update? Are we doing something wrong? Without this entitlement we cannot make the device ring in the background. The app is a voice and video messaging platform.

I am very new to the macOS Shortcuts application. In my opinion, the documentation is sparse and totally inadequate. The internet seems to be the only method of figuring out how to use this application. I have recently created a shortcut that is working well for me. It has several steps and is too large to fit in the Shortcuts editor window, so I cannot grab a screenshot of it for documenation purposes. I also cannot copy the contents of the editor window and paste them anywhere, such as in a new Note or TextEdit document. I do think Apple should add a means to create a PDF document of the Shortcuts editor window's contents. I went to the official Apple feedback page to leave comments but, irony of ironies, the Shortcuts app is not listed there! I have no idea what I am doing at this point but am excited to learn how to use Shortcuts to automate and simplify tasks that I have to perform frequently. Here's hoping the documentation and features for this application will one day be comprehensive, comprehensible, and complete.

Hi, I’m implementing a macOS DNS Proxy as a system extension and running into a persistent activation error: OSSystemExtensionErrorDomain error 9 (validationFailed) with the message: extension category returned error This happens both on an MDM‑managed Mac and on a completely clean Mac (no MDM, fresh install). Setup macOS: 15.x (clean machine, no MDM) Xcode: 16.x Team ID: AAAAAAA111 (test) Host app bundle ID: com.example.agent.NetShieldProxy DNS Proxy system extension bundle ID: com.example.agent.NetShieldProxy.dnsProxy The DNS Proxy is implemented as a NetworkExtension system extension, not an app extension. Host app entitlements From codesign -d --entitlements :- /Applications/NetShieldProxy.app: xml com.apple.application-identifier AAAAAAA111.com.example.agent.NetShieldProxy <key>com.apple.developer.system-extension.install</key> <true/> <key>com.apple.developer.team-identifier</key> <string>AAAAAAA111</string> <key>com.apple.security.app-sandbox</key> <true/> <key>com.apple.security.application-groups</key> <array> <string>group.com.example.NetShieldmac</string> </array> <key>com.apple.security.files.user-selected.read-only</key> <true/> xml com.apple.application-identifier AAAAAAA111.com.example.agent.NetShieldProxy.dnsProxy <key>com.apple.developer.networking.networkextension</key> <array> <string>dns-proxy-systemextension</string> </array> <key>com.apple.developer.team-identifier</key> <string>AAAAAAA111</string> <key>com.apple.security.application-groups</key> <array> <string>group.com.example.NetShieldmac</string> <string>group.example.NetShieldmac</string> <string>group.example.agent.enterprise.macos</string> <string>group.example.com.NetShieldmac</string> </array> DNS Proxy system extension Info.plist On the clean Mac, from: bash plutil -p "/Applications/NetShieldProxy.app/Contents/Library/SystemExtensions/com.example.agent.NetShieldProxy.dnsProxy.systemextension/Contents/Info.plist" I get: json { "CFBundleExecutable" => "com.example.agent.NetShieldProxy.dnsProxy", "CFBundleIdentifier" => "com.example.agent.NetShieldProxy.dnsProxy", "CFBundleName" => "com.example.agent.NetShieldProxy.dnsProxy", "CFBundlePackageType" => "SYSX", "CFBundleShortVersionString" => "1.0.1.8", "CFBundleSupportedPlatforms" => [ "MacOSX" ], "CFBundleVersion" => "0.1.1", "LSMinimumSystemVersion" => "13.5", "NSExtension" => { "NSExtensionPointIdentifier" => "com.apple.dns-proxy", "NSExtensionPrincipalClass" => "com_example_agent_NetShieldProxy_dnsProxy.DNSProxyProvider" }, "NSSystemExtensionUsageDescription" => "SYSTEM_EXTENSION_USAGE_DESCRIPTION" } The DNSProxyProvider class inherits from NEDNSProxyProvider and is built in the system extension target. Activation code In the host app, I use: swift import SystemExtensions final class SystemExtensionActivator: NSObject, OSSystemExtensionRequestDelegate { private let extensionIdentifier = "com.example.agent.NetShieldProxy.dnsProxy" func activate(completion: @escaping (Bool) -> Void) { let request = OSSystemExtensionRequest.activationRequest( forExtensionWithIdentifier: extensionIdentifier, queue: .main ) request.delegate = self OSSystemExtensionManager.shared.submitRequest(request) } func request(_ request: OSSystemExtensionRequest, didFailWithError error: Error) { let nsError = error as NSError print("Activation failed:", nsError) } func request(_ request: OSSystemExtensionRequest, didFinishWithResult result: OSSystemExtensionRequest.Result) { print("Result:", result.rawValue) } } Runtime behavior on a clean Mac (no MDM) config.plist is created under /Library/Application Support/NetShield (via a root shell script). A daemon runs, contacts our backend, and writes /Library/Application Support/NetShield/state.plist with a valid dnsToken and other fields. The app NetShieldProxy.app is installed via a notarized, stapled Developer ID .pkg. The extension bundle is present at: /Applications/NetShieldProxy.app/Contents/Library/SystemExtensions/com.example.agent.NetShieldProxy.dnsProxy.systemextension. When I press Activate DNS Proxy in the UI, I see in the unified log: text NetShieldProxy: [com.example.agent:SystemExtensionActivator] Requesting activation for system extension: com.example.agent.NetShieldProxy.dnsProxy NetShieldProxy: [com.example.agent:SystemExtensionActivator] SystemExtensionActivator - activation failed: extension category returned error (domain=OSSystemExtensionErrorDomain code=9) NetShieldProxy: [com.example.agent:SystemExtensionActivator] SystemExtensionActivator - OSSystemExtensionError code enum: 9 NetShieldProxy: [com.example.agent:SystemExtensionActivator] SystemExtensionActivator - validationFailed And: bash systemextensionsctl list -> 0 extension(s) There is no prompt in Privacy & Security on this clean Mac. Question Given: The extension is packaged as a system extension (CFBundlePackageType = SYSX) with NSExtensionPointIdentifier = "com.apple.dns-proxy". Host and extension share the same Team ID and Developer ID Application cert. Entitlements on the target machine match the provisioning profile and Apple’s docs for DNS Proxy system extensions (dns-proxy-systemextension). This is happening on a clean Mac with no MDM profiles at all. What are the likely reasons for OSSystemExtensionErrorDomain error 9 (validationFailed) with "extension category returned error" in this DNS Proxy system extension scenario? Is there any additional configuration required for DNS Proxy system extensions (beyond entitlements and Info.plist) that could trigger this category-level validation failure? Any guidance or examples of a working DNS Proxy system extension configuration (host entitlements + extension Info.plist + entitlements) would be greatly appreciated. Thanks!

I'm developing an iOS app that needs to continuously inform a server whether the user's paired Apple Watch is currently reachable for interactive messaging. If this reachability is lost unexpectedly, the server should be alerted within seconds. This is a safety-critical feature where reliability is essential. The goal (abstractly): The iPhone app needs real-time or near-real-time awareness of whether the paired Apple Watch is reachable. The specific mechanism doesn't matter - I'm open to any approach that achieves this reliably. Context - what already works: The iPhone app successfully maintains continuous server connectivity using an NEAppPushProvider network extension. In practice, this runs reliably in the background and sends periodic heartbeats to the server regardless of main app state. This pattern works well for the phone component. I need to extend this to include the watch's connectivity status in those server updates. Note: WCSession APIs are only available in the main app process, not the Network Extension, so any watch connectivity information must be bridged via the main iOS app (e.g. shared UserDefaults and Darwin notifications). What I've tried: 1. Companion watchOS app sending heartbeats to iPhone via WCSession This was my primary approach: a watchOS app sends messages to the iPhone at short intervals using WCSession.sendMessage(). The iPhone forwards this to the server. If heartbeats stop, the server raises an alert. (I tested various intervals from 2-15 seconds; the specific interval doesn't matter because the fundamental problem is that the watch app is suspended regardless.) Problem: The watch app is suspended almost immediately when: The user presses the Digital Crown The user switches to another app The watch screen dims and shows the clock face (even without explicit backgrounding) Once suspended, Timer.scheduledTimer() stops firing and no heartbeats are sent. 2. WCSession.isReachable monitoring on iPhone I hoped the iPhone could monitor WCSession.isReachable to detect when the watch becomes unreachable. Problem: isReachable indicates whether the counterpart app is reachable for interactive messaging, not the underlying physical connection. It returns false for many reasons - watch app suspended, backgrounded, or various system conditions - making it unreliable as a proxy for actual watch connectivity. The iPhone cannot distinguish "watch app not ready for messaging" from "watch physically disconnected". 3. WKExtendedRuntimeSession on watchOS Problem: Only available for specific scenarios (workout, mindfulness, etc.). My use case is general activity, not fitness tracking. Misusing workout sessions would likely be rejected by App Review. 4. WKApplicationRefreshBackgroundTask on watchOS Problem: These tasks are system-scheduled with timing that varies from minutes to hours depending on system conditions. Far too slow and unpredictable for second-level detection. 5. BLE advertising from watchOS app Problem: BLE advertising stops when the watchOS app is suspended. Same fundamental limitation as the timer approach. 6. Server directly pinging the watch (ICMP or similar) Problem: While Apple Watch can have an IP address via Wi-Fi or cellular (on LTE models), inbound connections to the watch aren't feasible - the watch is behind NAT with no public address, and watchOS doesn't support inbound server sockets (especially in background). This approach isn't practical regardless of connection type. 7. CoreBluetooth scanning from iPhone Problem: Apple Watch doesn't advertise as a discoverable BLE peripheral to third-party apps. The system-level pairing isn't exposed. Why this works on Android/WearOS: On WearOS, a Foreground Service continues running in the background regardless of UI state or screen status (subject to standard OS background limits, but in practice it works reliably). The service sends heartbeats via MessageClient consistently. This "always-on background execution" pattern has no equivalent on watchOS. Questions: Is there any mechanism for an iPhone app to have continuous or regularly-updated knowledge of whether a paired Apple Watch is connected and reachable for interactive messaging - ideally without requiring a watchOS companion app to be in the foreground? Are there any system-level APIs or entitlements (perhaps requiring special approval) that expose watch pairing/connectivity events to iOS apps? Is there any watchOS background execution mechanism I've missed that could keep code running reliably when the app isn't in the foreground? Has anyone solved a similar "detect wearable connectivity loss in real-time" problem on the Apple platform? I understand Apple designed watchOS with aggressive power management for good reasons. If continuous connectivity monitoring truly isn't possible, I'd appreciate confirmation so I can set appropriate user expectations. But given this is a safety-critical use case, I'm hoping there's an approach I've overlooked.

Hello, A customer has requested the development of a home assistance app to be published on the App Store. The app will connect to a server running locally at the end user's home, for example on a Raspberry Pi. Users would enter the IP address or hostname of their personal server into the app. A strict requirement is that, for data protection reasons, there must not be any proxy server. The app should only communicate directly with the local server (e.g., Raspberry Pi). We are able to solve technical challenges such as DNS, dynamic IP, and port forwarding, router configuration. However, I'm concerned about Apple's requirement that the endpoint – in this case, the Raspberry Pi at the user's home – must not use self-signed SSL certificates. While it may be technically possible to secure the home server with a certificate provider like Let's Encrypt, it is unrealistic to expect a typical user with no technical training to accomplish this setup independently. Is there a recommended solution to this problem, particularly in the context of IoT devices and apps? Any advice or experiences would be deeply appreciated.

I built an iOS app and debugged it using my iPhone 11. It works fine. My app uses Bluetooth because the physical data logger reads data via Bluetooth. I published it to the Apple store. After installing it on the iPhone 13 pro. The app works, but the device is not selected.

Issue Description: On non-Dynamic Island devices, when the screen is on, tapping the alarm banner does not open the app or trigger any action. Steps to Reproduce: Set a regular alarm. Wait until the alarm goes off. Keep the screen on. A banner appears with options (e.g., dismiss or a secondary action). Tap anywhere on the banner area that would normally open the app. Expected Behavior: Tapping the banner should open the app. Actual Behavior: Tapping the banner has no response.

Issue Description: When the snooze alarm and a set alarm share the same time, the behavior differs between locked and unlocked screen states. The current issue occurs when the screen is unlocked and the device is on the home screen before the alarm goes off: Alarm A is set for 17:23, and the snooze button is tapped when it rings. Alarm B is set for 17:25. At 17:25, Alarm B first vibrates and then rings (no buttons are pressed at this time). A few seconds later, it vibrates a second time. After that, the alarm becomes silent. If dynamic/notification alarms are disabled, the next scheduled alarm rings normally.

I am adding NSFileProviderPartialContentFetching support to an existing NSFileProviderReplicatedExtension. My backend has a high "Time To First Byte" latency (approx. 3 seconds) but reasonable throughput once the connection is established. I am observing a critical behavior difference between Partial Content Fetching and standard Materialization that causes sequential reads (e.g., dd, Finder copies, Adobe apps) to fail with timeouts. The Scenario: I have a 2.8 GB file. I attempt to read it sequentially using dd. **Baseline (Working): Partial Fetching Disabled ** I do not conform to NSFileProviderPartialContentFetching. The system triggers fetchContents(for:version:request:completionHandler:). My extension takes 3 seconds to connect, then streams the entire 2.8 GB file. Result: Success. The OS waits patiently for the entire download (minutes) without timing out, then dd reads the file instantly from the local disk. **The Issue: Partial Fetching Enabled ** I add conformance to NSFileProviderPartialContentFetching. The system requests small, aligned chunks (e.g., 16KB or 128KB). My extension fetches the requested range. This takes ~3 seconds due to network latency. The first few chunks succeed, but shortly after, the operation fails with Operation timed out. It appears the VFS kernel watchdog treats these repeated 3-second delays during read() syscalls as a stalled drive and kills the operation. **My Questions: ** Is there a documented timeout limit for fetchPartialContents completion handlers? It seems strictly enforced (similar to a local disk I/O timeout) compared to the lenient timeout for full materialization. Is NSFileProviderPartialContentFetching inherently unsuitable for high-latency backends (e.g., cold storage, slow handshakes), or is there a mechanism to signal "progress" to the kernel to reset the I/O watchdog during a slow partial fetch? Does the system treat partial fetching as "Online/Direct I/O" (blocking the user application) whereas full fetch is treated as "Offline/Syncing" (pausing the application), explaining the difference in tolerance? Any insights into the VFS lifecycle differences between these two modes would be appreciated.

Issue Description: Tapping the app widget sometimes triggers the Universal Link twice, causing duplicate navigation or actions within the app Steps to Reproduce: Add the app widget to the home screen Tap the widget to open the app via the Universal Link Observe that the Universal Link is sometimes fired twice Expected Behavior: Tapping the widget should trigger the Universal Link only once. Actual Behavior: Universal Link is triggered twice, causing duplicate navigation or actions.

Issue Description: When an alarm is set for a time earlier than the current system time, and the system time is manually adjusted back to before the alarm time, the alarm does not ring when the scheduled time is reached. Steps to Reproduce: Current time is 23:34 Set an alarm for 23:30 (earlier than the current time) Manually change the system time to 23:28 Wait until the time reaches 23:30 The alarm does not ring Device: iPhone 14 Pro / iOS 26.2

I have a Quick Action which flattens a PDF via Applescript. The code works extremely well--I right click in finder and the PDF is flattened, annotations are burned in, no other applications are opened, and the action completes in less than 2 seconds. Here is the Applescript code: I have a Shortcut which completes several operations on already-flattened PDFs. Presently 1) I run my Quick Action by right-clicking a PDF in Finder in order to flatten it, and 2) then right-click that save PDF in Finder to run my Shortcut on that now-flattened PDF. Ideally I'd like to add the Applescript code which flattens the PDF to the beginning of my Shortcut for the sake of convenience, and because sometimes I run my Shortcut having forgotten to flatten the PDF first. However I'm finding that the Applescript code, when placed into a Run Applescript action in Shortcuts, gives this error message: The exact same code when placed into a Run Applescript action in Automator and then run as a Quick Action by right-clicking the PDF in Finder, does not give an error and works perfectly. Does anybody have an explanation (and possible solution) for why this is the case?

I'm implementing PurchaseIntent.intents for App Store in-app purchase promotions, following Apple's WWDC guidance. The API only works on cold launch (killed→launch), but fails on background→foreground transitions, making App Store promotions unusable. Sample code as followed from WWDC23 video "What's new in StoreKit 2 and StoreKit Testing in Xcode". In the StoreKitManager observable class, I have this function which is initialized in a listening task: func listenForPurchaseIntent() -> Task<Void, Error> { return Task { [weak self] in for await purchase in PurchaseIntent.intents { guard let self else { continue } let product = purchase.product await self.purchaseProduct(product) } } } where purchaseProduct() will perform the call to: try await product.purchase() ISSUE: When the app is in background (after previously launched), and the purchase intent is initiated from Xcode Transaction Manager or using the "itms-services://?action=purchaseIntent" method, the system foregrounds my app but the purchase intent is never delivered to the waiting listener. The intent remains queued until the next cold launch (quit app and relaunch app). This could mean that if a user has installed the app, and has run the app, then tapped the promotional IAP from the App Store, the purchase intent will not show up until the next cold launch. If the app is in quit state, then the system will foreground the app, and purchase intent is delivered correctly. STEPS TO REPRODUCE Launch app (listener starts in StoreKitManager.init()) Background app Add purchase intent via Xcode Transaction Manager Foreground app Result: No purchase sheet appears, no intent delivered Workaround attempts: Using this either in a view or the main app: func checkForPurchaseIntents() async { for await purchaseIntent in PurchaseIntent.intents { await storeKit.purchaseProduct(purchaseIntent.product) } } Applied to .onChange(of: scenePhase) - Doesn't work, nothing happens. Using UIApplication.willEnterForegroundNotification - Only works on the first time the app goes from background to foreground when purchase intent is sent. Doesn't work on second time or third time. • Attempting to creating fresh listening task on each foreground - Does not work. The question is: How are we supposed to implement the PurchaseIntent API? I have checked Apple sample projects like BackyardBirds, and sample projects from WWDC on StoreKit 2 but they never implemented Purchase Intent.

DeviceActivityReportExtension issues Hi, I am currently working on a project and hoping to use DeviceActivityReportExtension. I have already successfully set up the DeviceActivityMonitorExtension and that is working correctly, but can't get this to reproduce. I initially had a complex capacitor project which displayed a native UI screen which (tried to) display the screen time report though it was failing silently. I tried a much simpler setup with a pure swift project (not an iOS dev so forgive my poor code quality) and still having an issue getting this to render on my computer. Here is a reproduction: https://github.com/ethanmichel0/screen-time-report-bug-recreation/tree/main Does this code work for others? and if not do you know how I can set this up? idk if my Xcode or iOS versions are incompatible (see versions in that link). Would be super grateful for some help, thanks so much.

The Problem The Family Activity Picker shows only the child's app categories on the guardian's/parent's device. The application names from the child's device are not showing on the guardian's/parent's device. The authorization is done on the child's device via try await AuthorizationCenter.shared.requestAuthorization(for: .child) Usage of the family activity picker on the guardian's/parent's device struct ContentView: View { @State private var isPresented = true @StateObject private var familyControlsHelper = FamilyControlsHelper.shared var onClose: () -> Void var body: some View { ZStack { Color.black.opacity(0.1).ignoresSafeArea() } .familyActivityPicker( isPresented: $isPresented, selection: $familyControlsHelper.familyActivitySelection ) .onChange(of: isPresented) { _ in if !isPresented { onClose() } } } } IMPORTANT Both devices are real (not simulators), and the app has granted distribution Family Controls entitlement. Question Is this the expected behavior? Or the child's app should appear on the guardian's device? Thanks.

So I'm aware that Apple can designate a folder as a "data vault", and access to that folder is limited to applications that have a specific entitlement. I was wondering if there was an equivalent (or the same, I'm not fussy :) feature available to third parties, even if only during the app-store submission ? To avoid the X-Y problem, what I want to do is have a launch agent with access to a SQLite database, and I want only that launch agent to have access. Any reads of the database will have to be done through an XPC call from the main user-facing application. I want to store private data into that database, and I don't want there to be any way for any other application to read it. If there's a way to do that without data-vaults I'm all ears :) I'm not sure if this is really the right place, perhaps the core-os forum would be better, but since the Apple solution is gate-kept by entitlements, I thought I'd start here :)

Hello Apple Bluetooth team, We are developing a real-time call translation system that streams raw PCM audio over BLE between iPhone and custom earbuds. This works reliably on iPhone 14 / 15 / 16, but on iPhone 17 (Bluetooth 6, N1 chip) we see severe and repeatable BLE packet loss, affecting both microphone uplink and TTS downlink. Our audio stream 16 kHz, 16-bit mono PCM 20 ms frames (~640 bytes) continuous bidirectional BLE streaming What happens on iPhone 17 BLE packets are frequently dropped entire audio frames are missing results in ASR gaps and broken TTS playback occurs even with strong RSSI and no RF interference Same firmware, same BLE protocol, same MTU and connection interval work normally on older iPhones. Questions We would like to know: Did Bluetooth 6 / N1 change BLE throughput, buffering, or scheduling? Are there new limits on sustained notify / write-without-response traffic? Is BLE audio now arbitrated differently against Wi-Fi / A2DP on iPhone 17? Is BLE still expected to support low-latency continuous audio streaming on iPhone 17, or is this no longer a safe assumption? Any guidance or new best practices would be greatly appreciated. Best regards, Valenti Zhang

Hi, I’m developing a Matter commissioning flow and would like to clarify Apple Home’s support for concatenated (multi-device) QR codes. In my implementation, I generate a single QR code that contains multiple Matter onboarding payloads (concatenated payloads), intended to commission multiple devices in one scan, similar to a multi-pack / multi-accessory flow. What I’ve tested: Standard single-device Matter QR codes work as expected in the Apple Home app A concatenated QR code (multiple Matter payloads combined into one QR) does not get recognized / commissioned by Apple Home My questions: Does Apple Home officially support commissioning via concatenated or multi-device Matter QR codes? If yes, is there a specific payload format or delimiter that Apple Home expects? If not, is this a known limitation or something planned for future iOS/Home releases?

I'm using StoreKit 2 with Product.products(for:) to fetch my auto-renewable subscriptions. It works in the Xcode simulator with a local StoreKit configuration file, but returns an empty array (no error) in TestFlight. iOS 15+, using async/await API Products are configured in App Store Connect Paid Apps agreement is active Sandbox tester account set up Has anyone experienced this? What am I missing?