My app Mocawave is a music player distributed through the Mac App Store. It declares specific audio document types (public.mp3, com.microsoft.waveform-audio, public.mpeg-4-audio, public.aac-audio) in its CFBundleDocumentTypes with a Viewer role. When a user sets Mocawave as the default app for audio files and double-clicks an MP3 downloaded from the internet (which has the com.apple.quarantine extended attribute), macOS displays the alert: "Apple could not verify [filename] is free of malware that may harm your Mac or compromise your privacy." This does not happen when: Opening the same file via NSOpenPanel from within the app Opening the same file with Apple's Music.app or QuickTime Player The app is: Distributed through the Mac App Store Sandboxed (com.apple.security.app-sandbox) Uses com.apple.security.files.user-selected.read-write entitlement The file being opened is a regular audio file (MP3), not an executable. Since the app is sandboxed and distributed through the App Store, I expected it to have sufficient trust to open quarantined data files without triggering Gatekeeper warnings — similar to how Music.app and QuickTime handle them. Questions: Is there a specific entitlement or Info.plist configuration that allows a sandboxed Mac App Store app to open quarantined audio files without this alert? Is this expected behavior for third-party App Store apps, or could this indicate a misconfiguration on my end? Environment: macOS 15 (Sequoia), app built with Swift/SwiftUI, targeting macOS 13+.

We are using SecItemCopyMatching from LocalAuthentication to access the private key to sign a challenge in our native iOS app twice in a few seconds from user interactions. This was working as expected up until about a week ago where we started getting reports of it hanging on the biometrics screen (see screenshot below). From our investigation we've found the following: It impacts newer iPhones using iOS 26.1 and later. We have replicated on these devices: iPhone 17 Pro max iPhone 16 Pro iPhone 15 Pro max iPhone 15 Only reproducible if the app tries to access the private key twice in quick succession after granting access to face ID. Looks like a race condition between the biometrics permission prompt and Keychain private key access We were able to make it work by waiting 10 seconds between private key actions, but this is terrible UX. We tried adding adding retries over the span of 10 seconds which fixed it on some devices, but not all. We checked the release notes for iOS 26.1, but there is nothing related to this. Screenshot:

% curl -v https://app-site-association.cdn-apple.com/a/v1/zfcs.bankts.cn Host app-site-association.cdn-apple.com:443 was resolved. IPv6: (none) IPv4: 218.92.226.151, 119.101.148.193, 218.92.226.6, 115.152.217.3 Trying 218.92.226.151:443... Connected to app-site-association.cdn-apple.com (218.92.226.151) port 443 ALPN: curl offers h2,http/1.1 (304) (OUT), TLS handshake, Client hello (1): CAfile: /etc/ssl/cert.pem CApath: none (304) (IN), TLS handshake, Server hello (2): (304) (IN), TLS handshake, Unknown (8): (304) (IN), TLS handshake, Certificate (11): (304) (IN), TLS handshake, CERT verify (15): (304) (IN), TLS handshake, Finished (20): (304) (OUT), TLS handshake, Finished (20): SSL connection using TLSv1.3 / AEAD-AES256-GCM-SHA384 / [blank] / UNDEF ALPN: server accepted http/1.1 Server certificate: subject: C=US; ST=California; O=Apple Inc.; CN=app-site-association.cdn-apple.com start date: Sep 25 13:58:08 2025 GMT expire date: Mar 31 17:44:25 2026 GMT subjectAltName: host "app-site-association.cdn-apple.com" matched cert's "app-site-association.cdn-apple.com" issuer: CN=Apple Public Server RSA CA 11 - G1; O=Apple Inc.; ST=California; C=US SSL certificate verify ok. using HTTP/1.x GET /a/v1/zfcs.bankts.cn HTTP/1.1 Host: app-site-association.cdn-apple.com User-Agent: curl/8.7.1 Accept: / Request completely sent off < HTTP/1.1 404 Not Found < Content-Type: text/plain; charset=utf-8 < Content-Length: 10 < Connection: keep-alive < Server: nginx < Date: Wed, 04 Feb 2026 02:26:00 GMT < Expires: Wed, 04 Feb 2026 02:26:10 GMT < Age: 24 < Apple-Failure-Details: {"cause":"context deadline exceeded (Client.Timeout exceeded while awaiting headers)"} < Apple-Failure-Reason: SWCERR00301 Timeout < Apple-From: https://zfcs.bankts.cn/.well-known/apple-app-site-association < Apple-Try-Direct: true < Vary: Accept-Encoding < Via: https/1.1 jptyo12-3p-pst-003.ts.apple.com (acdn/3.16363), http/1.1 jptyo12-3p-pac-043.ts.apple.com (acdn/3.16363), https/1.1 jptyo12-3p-pfe-002.ts.apple.com (acdn/3.16363) < X-Cache: MISS KS-CLOUD < CDNUUID: 736dc646-57fb-43c9-aa0d-eedad3a534f8-1154605242 < x-link-via: yancmp83:443;xmmp02:443;fzct321:443; < x-b2f-cs-cache: no-cache < X-Cache-Status: MISS from KS-CLOUD-FZ-CT-321-35 < X-Cache-Status: MISS from KS-CLOUD-XM-MP-02-16 < X-Cache-Status: MISS from KS-CLOUD-YANC-MP-83-15 < X-KSC-Request-ID: c4a640c815640ee93c263a357ee919d6 < CDN-Server: KSFTF < X-Cdn-Request-ID: c4a640c815640ee93c263a357ee919d6 < Not Found Connection #0 to host app-site-association.cdn-apple.com left intact

I'm trying to use ASWebAuthenticationSession on macOS but there is a weird crash and I have no idea what to do. It looks like there is a main thread check in a framework code that I have no control over. Any help would be appreciated. Thank you in advance. The stack of crashed thread has no symbols, even for supposedly my code in OAuthClient.authenticate. macOS 15.4.1 (24E263) Xcode Version 16.3 (16E140) Thread 11: EXC_BREAKPOINT (code=1, subcode=0x10039bb04) Thread 12 Queue : com.apple.NSXPCConnection.m-user.com.apple.SafariLaunchAgent (serial) #0 0x0000000100b17b04 in _dispatch_assert_queue_fail () #1 0x0000000100b52834 in dispatch_assert_queue$V2.cold.1 () #2 0x0000000100b17a88 in dispatch_assert_queue () #3 0x000000027db5f3e8 in swift_task_isCurrentExecutorWithFlagsImpl () #4 0x00000001022c7754 in closure #1 in closure #1 in OAuthClient.authenticate() () #5 0x00000001022d0c98 in thunk for @escaping @callee_guaranteed (@in_guaranteed URL?, @guaranteed Error?) -&gt; () () #6 0x00000001c7215a34 in __102-[ASWebAuthenticationSession initWithURL:callback:usingEphemeralSession:jitEnabled:completionHandler:]_block_invoke () #7 0x00000001c72163d0 in -[ASWebAuthenticationSession _endSessionWithCallbackURL:error:] () #8 0x00000001c7215fc0 in __43-[ASWebAuthenticationSession _startDryRun:]_block_invoke_2 () #9 0x0000000194e315f4 in __invoking___ () #10 0x0000000194e31484 in -[NSInvocation invoke] () #11 0x00000001960fd644 in __NSXPCCONNECTION_IS_CALLING_OUT_TO_REPLY_BLOCK__ () #12 0x00000001960fbe40 in -[NSXPCConnection _decodeAndInvokeReplyBlockWithEvent:sequence:replyInfo:] () #13 0x00000001960fb798 in __88-[NSXPCConnection _sendInvocation:orArguments:count:methodSignature:selector:withProxy:]_block_invoke_3 () #14 0x0000000194a6ef18 in _xpc_connection_reply_callout () #15 0x0000000194a6ee08 in _xpc_connection_call_reply_async () #16 0x0000000100b3130c in _dispatch_client_callout3_a () #17 0x0000000100b362f8 in _dispatch_mach_msg_async_reply_invoke () #18 0x0000000100b1d3a8 in _dispatch_lane_serial_drain () #19 0x0000000100b1e46c in _dispatch_lane_invoke () #20 0x0000000100b2bfbc in _dispatch_root_queue_drain_deferred_wlh () #21 0x0000000100b2b414 in _dispatch_workloop_worker_thread () #22 0x0000000100c0379c in _pthread_wqthread () My code: @MainActor func authenticate() async throws { let authURL = api.authorizationURL( scopes: scopes, state: state, redirectURI: redirectURI ) let authorizationCodeURL: URL = try await withUnsafeThrowingContinuation { c in let session = ASWebAuthenticationSession(url: authURL, callback: .customScheme(redirectScheme)) { url, error in guard let url = url else { c.resume(throwing: error ?? Error.unknownError("Failed to get authorization code")) return } c.resume(returning: url) } session.presentationContextProvider = presentationContextProvider session.start() } let authorizationCode = try codeFromAuthorizationURL(authorizationCodeURL) (storedAccessToken, storedRefreshToken) = try await getTokens(authorizationCode: authorizationCode) } Here is disassembly of the crashed function. libdispatch.dylib`_dispatch_assert_queue_fail: 0x10067fa8c &lt;+0&gt;: pacibsp 0x10067fa90 &lt;+4&gt;: sub sp, sp, #0x50 0x10067fa94 &lt;+8&gt;: stp x20, x19, [sp, #0x30] 0x10067fa98 &lt;+12&gt;: stp x29, x30, [sp, #0x40] 0x10067fa9c &lt;+16&gt;: add x29, sp, #0x40 0x10067faa0 &lt;+20&gt;: adrp x8, 71 0x10067faa4 &lt;+24&gt;: add x8, x8, #0x951 ; "not " 0x10067faa8 &lt;+28&gt;: adrp x9, 70 0x10067faac &lt;+32&gt;: add x9, x9, #0x16b ; "" 0x10067fab0 &lt;+36&gt;: stur xzr, [x29, #-0x18] 0x10067fab4 &lt;+40&gt;: cmp w1, #0x0 0x10067fab8 &lt;+44&gt;: csel x8, x9, x8, ne 0x10067fabc &lt;+48&gt;: ldr x10, [x0, #0x48] 0x10067fac0 &lt;+52&gt;: cmp x10, #0x0 0x10067fac4 &lt;+56&gt;: csel x9, x9, x10, eq 0x10067fac8 &lt;+60&gt;: stp x9, x0, [sp, #0x10] 0x10067facc &lt;+64&gt;: adrp x9, 71 0x10067fad0 &lt;+68&gt;: add x9, x9, #0x920 ; "BUG IN CLIENT OF LIBDISPATCH: Assertion failed: " 0x10067fad4 &lt;+72&gt;: stp x9, x8, [sp] 0x10067fad8 &lt;+76&gt;: adrp x1, 71 0x10067fadc &lt;+80&gt;: add x1, x1, #0x8eb ; "%sBlock was %sexpected to execute on queue [%s (%p)]" 0x10067fae0 &lt;+84&gt;: sub x0, x29, #0x18 0x10067fae4 &lt;+88&gt;: bl 0x1006c258c ; symbol stub for: asprintf 0x10067fae8 &lt;+92&gt;: ldur x19, [x29, #-0x18] 0x10067faec &lt;+96&gt;: str x19, [sp] 0x10067faf0 &lt;+100&gt;: adrp x0, 71 0x10067faf4 &lt;+104&gt;: add x0, x0, #0x956 ; "%s" 0x10067faf8 &lt;+108&gt;: bl 0x1006b7b64 ; _dispatch_log 0x10067fafc &lt;+112&gt;: adrp x8, 108 0x10067fb00 &lt;+116&gt;: str x19, [x8, #0x2a8] -&gt; 0x10067fb04 &lt;+120&gt;: brk #0x1

Estou compartilhando algumas observações técnicas sobre Crash Detection / Emergency SOS no ecossistema Apple, com base em eventos amplamente observados em 2022 e 2024, quando houve chamadas automáticas em massa para serviços de emergência. A ideia aqui não é discutir UX superficial ou “edge cases isolados”, mas sim comportamento sistêmico em escala, algo que acredito ser relevante para qualquer time que trabalhe com sistemas críticos orientados a eventos físicos. Contexto resumido A partir do iPhone 14, a Detecção de Acidente passou a correlacionar múltiplos sensores (acelerômetros de alta faixa, giroscópio, GPS, microfones) para inferir eventos de impacto severo e acionar automaticamente chamadas de emergência. Em 2022, isso resultou em um volume significativo de falsos positivos, especialmente em atividades com alta aceleração (esqui, snowboard, parques de diversão). Em 2024, apesar de ajustes, houve recorrência localizada do mesmo padrão. Ponto técnico central O problema não parece ser hardware, nem um “bug pontual”, mas sim o estado intermediário de decisão: Aceleração ≠ acidente Ruído ≠ impacto real Movimento extremo ≠ incapacidade humana Quando o classificador entra em estado ambíguo, o sistema depende de uma janela curta de confirmação humana (toque/voz). Em ambientes ruidosos, com o usuário em movimento ou fisicamente ativo, essa confirmação frequentemente falha. O sistema então assume incapacidade e executa a ação fail-safe: chamada automática. Do ponto de vista de engenharia de segurança, isso é compreensível. Do ponto de vista de escala, é explosivo. Papel da Siri A Siri não “decide” o acidente, mas é um elo sensível na cadeia humano–máquina. Falhas de compreensão por ruído, idioma, respiração ofegante ou ausência de resposta acabam sendo interpretadas como sinal de emergência real. Isso é funcionalmente equivalente ao que vemos em sistemas automotivos como o eCall europeu, quando a confirmação humana é inexistente ou degradada. O dilema estrutural Há um trade-off claro e inevitável: Reduzir falsos negativos (não perder um acidente real) Aumentar falsos positivos (chamadas indevidas) Para o usuário individual, errar “para mais” faz sentido. Para serviços públicos de emergência, milhões de dispositivos errando “para mais” criam ruído operacional real. Por que isso importa para developers A Apple hoje opera, na prática, um dos maiores sistemas privados de segurança pessoal automatizada do mundo, interagindo diretamente com infraestrutura pública crítica. Isso coloca Crash Detection / SOS na mesma categoria de sistemas safety-critical, onde: UX é parte da segurança Algoritmos precisam ser auditáveis “Human-in-the-loop” não pode ser apenas nominal Reflexões abertas Alguns pontos que, como developer, acho que merecem discussão: Janelas de confirmação humana adaptativas ao contexto (atividade física, ruído). Cancelamento visual mais agressivo em cenários de alto movimento. Perfis de sensibilidade por tipo de atividade, claramente comunicados. Critérios adicionais antes da chamada automática quando o risco de falso positivo é estatisticamente alto. Não é um problema simples, nem exclusivo da Apple. É um problema de software crítico em contato direto com o mundo físico, operando em escala planetária. Justamente por isso, acho que vale uma discussão técnica aberta, sem ruído emocional. Curioso para ouvir perspectivas de quem trabalha com sistemas similares (automotivo, wearables, safety-critical, ML embarcado). — Rafa

I'm experiencing a strange issue where ASWebAuthenticationSession works perfectly when running from Xcode (both Debug and Release), but fails on TestFlight builds. The setup: iOS app using ASWebAuthenticationSession for OIDC login (Keycloak) Custom URL scheme callback (myapp://) prefersEphemeralWebBrowserSession = false The issue: When using iOS Keychain autofill (with Face ID/Touch ID or normal iphone pw, that auto-submits the form) -> works perfectly When manually typing credentials and clicking the login button -> fails with white screen When it fails, the form POST from Keycloak back to my server (/signin-oidc) never reaches the server at all. The authentication session just shows a white screen. Reproduced on: Multiple devices (iPhone 15 Pro, etc.) iOS 18.x Xcode 16.x Multiple TestFlight testers confirmed same behavior What I've tried: Clearing Safari cookies/data prefersEphemeralWebBrowserSession = true and false Different SameSite cookie policies on server Verified custom URL scheme is registered and works (testing myapp://test in Safari opens the app) Why custom URL scheme instead of Universal Links: We couldn't get Universal Links to trigger from a js redirect (window.location.href) within ASWebAuthenticationSession. Only custom URL schemes seemed to be intercepted. If there's a way to make Universal Links work in this context, without a manual user-interaction we'd be happy to try. iOS Keychain autofill works The only working path is iOS Keychain autofill that requires iphone-authentication and auto-submits the form. Any manual form submission fails, but only on TestFlight - not Xcode builds. Has anyone encountered this or know a workaround?

We are using Apple's PSSO to federate device login to out own IdP. We have developed our own extension app and deployed it using MDM. Things works fine but there are 2 issues that we are trying to get to the root cause - On some devices after restarting we see an error message on the logic screen saying "The registration for this device is invalid and must be repaired" And other error message is "SmartCard configuration is invalid for this account" For the 1st we have figured out that this happens when the registration doesn't happen fully and the key is not tied to the user so when the disk needs to be decrypted at the FileVault screen the issue is raised. For the "SmartCard configuration is invalid for this account" issue also one aspect is invalid registration but there has been other instances as well where the devices were registered completely but then also the the above error was raised. We verified the registration being completed by checking if the SmartCard is visible in the System Report containing the key. Has anyone seen the above issues and any possible resolution around it?

Trusted execution is a generic name for a Gatekeeper and other technologies that aim to protect users from malicious code. General: Forums topic: Code Signing Forums tag: Gatekeeper Developer > Signing Mac Software with Developer ID Apple Platform Security support document Safely open apps on your Mac support article Hardened Runtime document WWDC 2022 Session 10096 What’s new in privacy covers some important Gatekeeper changes in macOS 13 (starting at 04: 32), most notably app bundle protection WWDC 2023 Session 10053 What’s new in privacy covers an important change in macOS 14 (starting at 17:46), namely, app container protection WWDC 2024 Session 10123 What’s new in privacy covers an important change in macOS 15 (starting at 12:23), namely, app group container protection Updates to runtime protection in macOS Sequoia news post Testing a Notarised Product forums post Resolving Trusted Execution Problems forums post App Translocation Notes (aka Gatekeeper path randomisation) forums post Most trusted execution problems are caused by code signing or notarisation issues. See Code Signing Resources and Notarisation Resources. Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = "eskimo" + "1" + "@" + "apple.com"

I regularly see folks confused by the difference in behaviour of app groups between macOS and iOS. There have been substantial changes in this space recently. While much of this is now covered in the official docs (r. 92322409), I’ve updated this post to go into all the gory details. If you have questions or comments, start a new thread with the details. Put it in the App & System Services > Core OS topic area and tag it with Code Signing and Entitlements. Oh, and if your question is about app group containers, also include Files and Storage. Share and Enjoy — Quinn “The Eskimo!” @ Developer Technical Support @ Apple let myEmail = "eskimo" + "1" + "@" + "apple.com" App Groups: macOS vs iOS: Working Towards Harmony There are two styles of app group ID: iOS-style app group IDs start with group., for example, group.eskimo1.test. macOS-style app group IDs start with your Team ID, for example, SKMME9E2Y8.eskimo1.test. This difference has been the source of numerous weird problems over the years. Starting in Feb 2025, iOS-style app group IDs are fully supported on macOS for all product types [1]. If you’re writing new code that uses app groups, use an iOS-style app group ID. If you have existing code that uses a macOS-style app group ID, consider how you might transition to the iOS style. IMPORTANT The Feb 2025 changes aren’t tied to an OS release but rather to a Developer website update. For more on this, see Feb 2025 Changes, below. [1] If your product is a standalone executable, like a daemon or agent, wrap it in an app-like structure, as explained in Signing a daemon with a restricted entitlement. iOS-Style App Group IDs An iOS-style app group ID has the following features: It starts with the group. prefix, for example, group.eskimo1.test. You allocate it on the Developer website. This assigns the app group ID to your team. You then claim access to it by listing it in the App Groups entitlement (com.apple.security.application-groups) entitlement. That claim must be authorised by a provisioning profile [1]. The Developer website will only let you include your team’s app group IDs in your profile. For more background on provisioning profiles, see TN3125 Inside Code Signing: Provisioning Profiles. iOS-style app group IDs originated on iOS with iOS 3.0. They’ve always been supported on iOS’s child platforms (iPadOS, tvOS, visionOS, and watchOS). On the Mac: They’ve been supported by Mac Catalyst since that technology was introduced. Likewise for iOS Apps on Mac. Starting in Feb 2025, they’re supported for other Mac products. [1] Strictly speaking macOS does not require that, but if your claim is not authorised by a profile then you might run into other problems. See Entitlements-Validated Flag, below. macOS-Style App Group IDs A macOS-style app group ID has the following features: It should start with your Team ID [1], for example, SKMME9E2Y8.eskimo1.test. It can’t be explicitly allocated on the Developer website. Code that isn’t sandboxed doesn’t need to claim the app group ID in the App Groups entitlement. [2] To use an app group, claim the app group ID in the App Groups entitlement. The App Groups entitlement is not restricted on macOS, meaning that this claim doesn’t need to be authorised by a provisioning profile [3]. However, if you claim an app group ID that’s not authorised in some way, you might run into problems. More on that later in this post. If you submit an app to the Mac App Store, the submission process checks that your app group IDs make sense, that is, they either start with your Team ID (macOS style) or are assigned to your team (iOS style). [1] This is “should” because, historically, macOS has not actually required it. However, that’s now changing, with things like app group container protection. [2] This was true prior to macOS 15. It may still technically be true in macOS 15 and later, but the most important thing, access to the app group container, requires the entitlement because of app group container protection. [3] Technically it’s a validation-required entitlement, something that we’ll come back to in the Entitlements-Validated Flag section. Feb 2025 Changes On 21 Feb 2025 we rolled out a change to the Developer website that completes the support for iOS-style app group IDs on the Mac. Specifically, it’s now possible to create a Mac provisioning profile that authorises the use of an iOS-style app group ID. Note This change doesn’t affect Mac Catalyst or iOS Apps on Mac, which have always been able to use iOS-style app group IDs on the Mac. Prior to this change it was possible to use an iOS-style app group ID on the Mac but that might result in some weird behaviour. Later sections of this post describe some of those problems. Of course, that information is now only of historical interest because, if you’re using an iOS-style app group, you can and should authorise that use with a provisioning profile. We also started seeding Xcode 16.3, which has since been release. This is aware of the Developer website change, and its Signing & Capabilities editor actively encourages you to use iOS-style app groups IDs in all products. Note This Xcode behaviour is the only option for iOS and its child platforms. With Xcode 16.3, it’s now the default for macOS as well. If you have existing project, enable this behaviour using the Register App Groups build setting. Finally, we updated a number of app group documentation pages, including App Groups entitlement and Configuring app groups. Crossing the Streams In some circumstances you might need to have a single app that accesses both an iOS- and a macOS-style app group. For example: You have a macOS app. You want to migrate to an iOS-style app group ID, perhaps because you want to share an app group container with a Mac Catalyst app. But you also need to access existing content in a container identified by a macOS-style app group ID. Historically this caused problems (FB16664827) but, as of Jun 2025, this is fully supported (r. 148552377). When the Developer website generates a Mac provisioning profile for an App ID with the App Groups capability, it automatically adds TEAM_ID.* to the list of app group IDs authorised by that profile (where TEAM_ID is your Team ID). This allows the app to claim access to every iOS-style app group ID associated with the App ID and any macOS-style app group IDs for that team. This helps in two circumstances: It avoids any Mac App Store Connect submission problems, because App Store Connect can see that the app’s profile authorises its use of all the it app group IDs it claims access to. Outside of App Store — for example, when you directly distribute an app using Developer ID signing — you no longer have to rely on macOS granting implicit access to macOS-style app group IDs. Rather, such access is explicitly authorised by your profile. That ensures that your entitlements remain validated, as discussed in the Entitlements-Validated Flag, below. A Historical Interlude These different styles of app group IDs have historical roots: On iOS, third-party apps have always used provisioning profiles, and thus the App Groups entitlement is restricted just like any other entitlement. On macOS, support for app groups was introduced before macOS had general support for provisioning profiles [1], and thus the App Groups entitlement is unrestricted. The unrestricted nature of this entitlement poses two problems. The first is accidental collisions. How do you prevent folks from accidentally using an app group ID that’s in use by some other developer? On iOS this is easy: The Developer website assigns each app group ID to a specific team, which guarantees uniqueness. macOS achieved a similar result by using the Team ID as a prefix. The second problem is malicious reuse. How do you prevent a Mac app from accessing the app group containers of some other team? Again, this isn’t an issue on iOS because the App Groups entitlement is restricted. On macOS the solution was for the Mac App Store to prevent you from publishing an app that used an app group ID that’s used by another team. However, this only works for Mac App Store apps. Directly distributed apps were free to access app group containers of any other app. That was considered acceptable back when the Mac App Store was first introduced. That’s no longer the case, which is why macOS 15 introduced app group container protection. See App Group Container Protection, below. [1] I’m specifically talking about provisioning profiles for directly distributed apps, that is, apps using Developer ID signing. Entitlements-Validated Flag The fact that the App Groups entitlement is unrestricted on macOS is, when you think about it, a little odd. The purpose of entitlements is to gate access to functionality. If an entitlement isn’t restricted, it’s not much of a gate! For most unrestricted entitlements that’s not a problem. Specifically, for both the App Sandbox and Hardened Runtime entitlements, those are things you opt in to, so macOS is happy to accept the entitlement at face value. After all, if you want to cheat you can just not opt in [1]. However, this isn’t the case for the App Groups entitlement, which actually gates access to functionality. Dealing with this requires macOS to walk a fine line between security and compatibility. Part of that solution is the entitlements-validated flag. When a process runs an executable, macOS checks its entitlements. There are two categories: Restricted entitlements must be authorised by a provisioning profile. If your process runs an executable that claims a restricted entitlement that’s not authorised by a profile, the system traps. Unrestricted entitlements don’t have to be authorised by a provisioning profile; they can be used by any code at any time. However, the App Groups entitlement is a special type of unrestricted entitlement called a validation-required entitlement. If a process runs an executable that claims a validation-required entitlement and that claim is not authorised by a profile, the system allows the process to continue running but clears its entitlements-validated flag. Some subsystems gate functionality on the entitlements-validated flag. For example, the data protection keychain uses entitlements as part of its access control model, but refuses to honour those entitlements if the entitlement-validated flag has been cleared. Note If you’re curious about this flag, use the procinfo subcommand of launchctl to view it. For example: % sudo launchctl procinfo `pgrep Test20230126` … code signing info = valid … entitlements validated … If the flag has been cleared, this line will be missing from the code signing info section. Historically this was a serious problem because it prevented you from creating an app that uses both app groups and the data protection keychain [2] (r. 104859788). Fortunately that’s no longer an issue because the Developer website now lets you include the App Groups entitlement in macOS provisioning profiles. [1] From the perspective of macOS checking entitlements at runtime. There are other checks: The App Sandbox is mandatory for Mac App Store apps, but that’s checked when you upload the app to App Store Connect. Directly distributed apps must be notarised to pass Gatekeeper, and the notary service requires that all executables enable the hardened runtime. [2] See TN3137 On Mac keychain APIs and implementations for more about the data protection keychain. App Groups and the Keychain The differences described above explain a historical oddity associated with keychain access. The Sharing access to keychain items among a collection of apps article says: Application groups When you collect related apps into an application group using the App Groups entitlement, they share access to a group container, and gain the ability to message each other in certain ways. You can use app group names as keychain access group names, without adding them to the Keychain Access Groups entitlement. On iOS this makes a lot of sense: The App Groups entitlement is a restricted entitlement on iOS. The Developer website assigns each iOS-style app group ID to a specific team, which guarantees uniqueness. The required group. prefix means that these keychain access groups can’t collide with other keychain access groups, which all start with an App ID prefix (there’s also Apple-only keychain access groups that start with other prefixes, like apple). However, this didn’t work on macOS [1] because the App Groups entitlement is unrestricted there. However, with the Feb 2025 changes it should now be possible to use an iOS-style app group ID as a keychain access group on macOS. Note I say “should” because I’ve not actually tried it (-: Keep in mind that standard keychain access groups are protected the same way on all platforms, using the restricted Keychain Access Groups entitlement (keychain-access-groups). [1] Except for Mac Catalyst apps and iOS Apps on Mac. Not Entirely Unsatisfied When you launch a Mac app that uses app groups you might see this log entry: type: error time: 10:41:35.858009+0000 process: taskgated-helper subsystem: com.apple.ManagedClient category: ProvisioningProfiles message: com.example.apple-samplecode.Test92322409: Unsatisfied entitlements: com.apple.security.application-groups Note The exact format of that log entry, and the circumstances under which it’s generated, varies by platform. On macOS 13.0.1 I was able to generate it by running a sandboxed app that claims a macOS-style app group ID in the App Groups entitlement and also claims some other restricted entitlement. This looks kinda worrying and can be the source of problems. It means that the App Groups entitlement claims an entitlement that’s not authorised by a provisioning profile. On iOS this would trap, but on macOS the system allows the process to continue running. It does, however, clear the entitlements-validate flag. See Entitlements-Validated Flag for an in-depth discussion of this. The easiest way to avoid this problem is to authorise your app group ID claims with a provisioning profile. If there’s some reason you can’t do that, watch out for potential problems with: The data protection keychain — See the discussion of that in the Entitlements-Validated Flag and App Groups and the Keychain sections, both above. App group container protection — See App Group Container Protection, below. App Group Container Protection macOS 15 introduced app group container protection. To access an app group container without user intervention: Claim access to the app group by listing its ID in the App Groups entitlement. Locate the container by calling the containerURL(forSecurityApplicationGroupIdentifier:) method. Ensure that at least one of the following criteria are met: Your app is deployed via the Mac App Store (A). Or via TestFlight when running on macOS 15.1 or later (B). Or the app group ID starts with your app’s Team ID (C). Or your app’s claim to the app group is authorised by a provisioning profile embedded in the app (D) [1]. If your app doesn’t follow these rules, the system prompts the user to approve its access to the container. If granted, that consent applies only for the duration of that app instance. For more on this, see: The System Integrity Protection section of the macOS Sequoia 15 Release Notes The System Integrity Protection section of the macOS Sequoia 15.1 Release Notes WWDC 2024 Session 10123 What’s new in privacy, starting at 12:23 The above criteria mean that you rarely run into the app group authorisation prompt. If you encounter a case where that happens, feel free to start a thread here on DevForums. See the top of this post for info on the topic and tags to use. Note Prior to the Feb 2025 change, things generally worked out fine when you app was deployed but you might’ve run into problems during development. That’s no longer the case. [1] This is what allows Mac Catalyst and iOS Apps on Mac to work. Revision History 2025-08-12 Added a reference to the Register App Groups build setting. 2025-07-28 Updated the Crossing the Streams section for the Jun 2025 change. Made other minor editorial changes. 2025-04-16 Rewrote the document now that iOS-style app group IDs are fully supported on the Mac. Changed the title from App Groups: macOS vs iOS: Fight! to App Groups: macOS vs iOS: Working Towards Harmony 2025-02-25 Fixed the Xcode version number mentioned in yesterday’s update. 2025-02-24 Added a quick update about the iOS-style app group IDs on macOS issue. 2024-11-05 Further clarified app group container protection. Reworked some other sections to account for this new reality. 2024-10-29 Clarified the points in App Group Container Protection. 2024-10-23 Fleshed out the discussion of app group container protection on macOS 15. 2024-09-04 Added information about app group container protection on macOS 15. 2023-01-31 Renamed the Not Entirely Unsatisfactory section to Not Entirely Unsatisfied. Updated it to describe the real impact of that log message. 2022-12-12 First posted.

Hi everyone, I’d like to clarify something regarding the behavior of Team IDs after an app transfer between Apple Developer accounts. I have an app update that enforces a force update for all users. My plan is to release this update under the current developer account, and then proceed with transferring the app to a different developer account shortly afterward. My concern is: once the transfer is complete, will users who download the same app version (released before the transfer) be logged out due to a change in Team ID? Specifically, does the transferred app continue to use the original Team ID (used to sign the last submitted build), or does the Team ID change immediately upon transfer — affecting Keychain access? Any insights or confirmation on this would be greatly appreciated. Thanks!

I have a sandboxed Mac app which I can grant access to a folder using an NSOpenPanel. Once it’s been granted access it can enumerate the contents of the folder just fine. If I rename the folder while the app is open and then make the app enumerate the folder’s contents again, though, it seems to have lost access. What’s the recommended way to have an app’s sandbox “track” files as they’re moved around the filesystem? (NSDocument handles this for you, from what I can tell.) I’ve managed to hack something together with a combination of Dispatch sources and security-scoped bookmarks, but it feels like there must be an easier solution …

In the hopes of saving others time, the updated demo project (i.e. the new Shiny) can be found from the video 'Resources' section under 'Performing fast account creation with passkeys'. The beta documentation can also be found from there. All of the new functionality is available only on *OS 26 at this time.

The header documentation for the (deprecated) LAContext.evaluatedPolicyDomainState property contains the following: @warning Please note that the value returned by this property can change exceptionally between major OS versions even if the state of biometry has not changed. I noticed that the documentation for the new LAContext.domainState property does not contain a similar warning. I also found this related thread from 2016/17. Is the domainState property not susceptible to changes between major OS versions? Or is this generally not an issue anymore?

Hi everyone, I'm developing a C++ plugin (.bundle) for a third-party host application (Autodesk Maya) on macOS, and I'm finalizing the design for our licensing system. The plugin is distributed outside the Mac App Store. My goal is to securely store a license key in the user's Keychain. After some research, my proposed implementation is as follows: On activation, store the license data in the user's login keychain as a Generic Password (kSecClassGenericPassword) using the SecItem APIs. To ensure the plugin can access the item when loaded by Maya, I will use a specific Keychain Access Group (e.g., MY_TEAM_ID.com.mywebsite). The final .bundle will be code-signed with our company's Developer ID certificate. The signature will include an entitlements file (.entitlements) that specifies the matching keychain-access-groups permission. My understanding is that this combination of a unique Keychain Access Group and a properly signed/entitled bundle is the key to getting reliable Keychain access. This should also correctly trigger the one-time user permission prompt on first use. Does this sound like the correct and most robust approach for this scenario? Are there any common pitfalls with a plugin's Keychain access from within a host app that I should be aware of? Thanks for any feedback!

My high-level goal is to add support for Game Mode in a Java game, which launches via a macOS "launcher" app that runs the actual java game as a separate process (e.g. using the java command line tool). I asked this over in the Graphics & Games section and was told this, which is why I'm reposting this here. I'm uncertain how to speak to CLI tools and Java games launched from a macOS app. These sound like security and sandboxing questions which we recommend you ask about in those sections of the forums. The system seems to decide whether to enable Game Mode based on values in the Info.plist (e.g. for LSApplicationCategoryType and GCSupportsGameMode). However, the child process can't seem to see these values. Is there a way to change that? (The rest of this post is copied from my other forums post to provide additional context.) Imagine a native macOS app that acts as a "launcher" for a Java game.** For example, the "launcher" app might use the Swift Process API or a similar method to run the java command line tool (lets assume the user has installed Java themselves) to run the game. I have seen How to Enable Game Mode. If the native launcher app's Info.plist has the following keys set: LSApplicationCategoryType set to public.app-category.games LSSupportsGameMode set to true (for macOS 26+) GCSupportsGameMode set to true The launcher itself can cause Game Mode to activate if the launcher is fullscreened. However, if the launcher opens a Java process that opens a window, then the Java window is fullscreened, Game Mode doesn't seem to activate. In this case activating Game Mode for the launcher itself is unnecessary, but you'd expect Game Mode to activate when the actual game in the Java window is fullscreened. Is there a way to get Game Mode to activate in the latter case? ** The concrete case I'm thinking of is a third-party Minecraft Java Edition launcher, but the issue can also be demonstrated in a sample project (FB13786152). It seems like the official Minecraft launcher is able to do this, though it's not clear how. (Is its bundle identifier hardcoded in the OS to allow for this? Changing a sample app's bundle identifier to be the same as the official Minecraft launcher gets the behavior I want, but obviously this is not a practical solution.)

I'm seeing some odd behavior which may be a bug. I've broken it down to a least common denominator to reproduce it. But maybe I'm doing something wrong. I am opening a file read-write. I'm then mapping the file read-only and private: void* pointer = mmap(NULL, 17, PROT_READ, MAP_FILE | MAP_PRIVATE, fd, 0); I then unmap the memory and close the file. After the close, eslogger shows me this: {"close":{"modified":false,[...],"was_mapped_writable":false}} Which makes sense. I then change the mmap statement to: void* pointer = mmap(NULL, 17, PROT_READ, MAP_FILE | MAP_SHARED, fd, 0); I run the new code and and the close looks like: {"close":{"modified":false, [....], "was_mapped_writable":true}} Which also makes sense. I then run the original again (ie, with MAP_PRIVATE vs. MAP_SHARED) and the close looks like: {"close":{"modified":false,"was_mapped_writable":true,[...]} Which doesn't appear to be correct. Now if I just open and close the file (again, read-write) and don't mmap anything the close still shows: {"close":{ [...], "was_mapped_writable":true,"modified":false}} And the same is true if I open the file read-only. It will remain that way until I delete the file. If I recreate the file and try again, everything is good until I map it MAP_SHARED. I tried this with macOS 13.6.7 and macOS 15.0.1.

I work for Brave, a browser with ~80M users. We want to introduce a new system for automatic updates called Omaha 4 (O4). It's the same system that powers automatic updates in Chrome. O4 runs as a separate application on users' systems. For Chrome, this works as follows: An app called GoogleUpdater.app regularly checks for updates in the background. When a new version is found, then GoogleUpdater.app installs it into Chrome's installation directory /Applications/Google Chrome.app. But consider what this means: A separate application, GoogleUpdater.app, is able to modify Google Chrome.app. This is especially surprising because, for example, the built-in Terminal.app is not able to modify Google Chrome.app. Here's how you can check this for yourself: (Re-)install Chrome with its DMG installer. Run the following command in Terminal: mkdir /Applications/Google\ Chrome.app/test. This works. Undo the command: rm -rf /Applications/Google\ Chrome.app/test Start Chrome and close it again. mkdir /Applications/Google\ Chrome.app/test now fails with "Operation not permitted". (These steps assume that Terminal does not have Full Disk Access and System Integrity Protection is enabled.) In other words, once Chrome was started at least once, another application (Terminal in this case) is no longer allowed to modify it. But at the same time, GoogleUpdater.app is able to modify Chrome. It regularly applies updates to the browser. For each update, this process begins with an mkdir call similarly to the one shown above. How is this possible? What is it in macOS that lets GoogleUpdater.app modify Chrome, but not another app such as Terminal? Note that Terminal is not sandboxed. I've checked that it's not related to codesigning or notarization issues. In our case, the main application (Brave) and the updater (BraveUpdater) are signed and notarized with the same certificate and have equivalent requirements, entitlements and provisioning profiles as Chrome and GoogleUpdater. The error that shows up in the Console for the disallowed mkdir call is: kernel (Sandbox) System Policy: mkdir(8917) deny(1) file-write-create /Applications/Google Chrome.app/foo (It's a similar error when BraveUpdater tries to install a new version into /Applications/Brave Browser.app.) The error goes away when I disable System Integrity Protection. But of course, we cannot ask users to do that. Any help would be greatly appreciated.

Problem Statement: Pre-requisite is to generate a PKCS#12 file using openssl 3.x or above. Note: I have created a sample cert, but unable to upload it to this thread. Let me know if there is a different way I can upload. When trying to import a p12 certificate (generated using openssl 3.x) using SecPKCS12Import on MacOS (tried on Ventura, Sonoma, Sequoia). It is failing with the error code: -25264 and error message: MAC verification failed during PKCS12 import (wrong password?). I have tried importing in multiple ways through, Security Framework API (SecPKCS12Import) CLI (security import &lt;cert_name&gt; -k ~/Library/Keychains/login.keychain -P "&lt;password&gt;”) Drag and drop in to the Keychain Application All of them fail to import the p12 cert. RCA: The issues seems to be due to the difference in the MAC algorithm. The MAC algorithm used in the modern certs (by OpenSSL3 is SHA-256) which is not supported by the APPLE’s Security Framework. The keychain seems to be expecting the MAC algorithm to be SHA-1. Workaround: The current workaround is to convert the modern p12 cert to a legacy format (using openssl legacy provider which uses openssl 1.1.x consisting of insecure algorithms) which the SecPKCS12Import API understands. I have created a sample code using references from another similar thread (https://developer.apple.com/forums/thread/723242) from 2023. The steps to compile and execute the sample is mentioned in the same file. PFA the sample code by the name “pkcs12_modern_to_legacy_converter.cpp”. Also PFA a sample certificate which will help reproduce the issue by the name “modern_certificate.p12” whose password is “export”. Questions: Is there a fix on this issue? If yes, pls guide me through it; else, is it expected to be fixed in the future releases? Is there a different way to import the p12 cert which is resistant to the issue? This issue also poses a security concerns on using outdated cryptographic algorithms. Kindly share your thoughts. pkcs12_modern_to_legacy_converter.cpp

Hello, I want to access the Docker socket API from inside the macOS App Sandbox. The method queries the API using curl with --unix-socket. However, the Sandbox blocks the request, as shown by the log: curl(22299) deny(1) network-outbound /Users/user/.docker/run/docker.sock Outgoing network traffic is generally allowed, but access to the Docker Unix socket is denied. Here’s the code I’m using: private func executeDockerAPI() -> String { let process = Process() let pipe = Pipe() process.executableURL = URL(fileURLWithPath: "/usr/bin/curl") process.arguments = [ "--unix-socket", "/Users/user/.docker/run/docker.sock", "http://127.0.0.1/containers/json" ] process.standardOutput = pipe process.standardError = pipe do { try process.run() process.waitUntilExit() let data = pipe.fileHandleForReading.readDataToEndOfFile() if let output = String(data: data, encoding: .utf8) { return output } else { return "Error while decoding" } } catch { return "Error running command: \(error.localizedDescription)" } } Is there any entitlement or sandbox configuration I’m missing to allow access to /Users/user/.docker/run/docker.sock from inside the sandbox?

We have a macOS app that has a Photos Extension, which shares documents with the app via an app group container. Historically we used to have an iOS-style group identifier (group.${TeamIdentifier}${groupName}), because we were lead by the web interface in the developer portal to believe this to be the right way to name groups. Later with the first macOS 15 betas last year there was a bug with the operating system warning users, our app would access data from different apps, but it was our own app group container directory. Therefore we added a macOS-style group identifier (${TeamIdentifier}${groupName}) and wrote a migration of documents to the new group container directory. So basically we need to have access to these two app group containers for the foreseeable future. Now with the introduction of iOS-style group identifiers for macOS, Xcode Cloud no longer archives our app for TestFlight or AppStore, because it complains: ITMS-90286: Invalid code signing entitlements - Your application bundle’s signature contains code signing entitlements that aren’t supported on macOS. Specifically, the “[group.${TeamIdentifier}${groupName}, ${TeamIdentifier}${groupName}]” value for the com.apple.security.application-groups key in isn’t supported. This value should be a string or an array of strings, where each string is the “group” value or your Team ID, followed by a dot (“.”), followed by the group name. If you're using the “group” prefix, verify that the provisioning profile used to sign the app contains the com.apple.security.application-groups entitlement and its associated value(s). We have included the iOS-style group identifier in the provisioning profile, generated automatically, but can't do the same for the macOS-style group identifier, because the web interface only accepts identifiers starting with "group". How can we get Xcode Cloud to archive our app again using both group identifiers? Thanks in advance