feat(macos): per-app audio selection for screen sharing

Let users choose what the SCREEN_AUDIO stream captures before sharing:
share everything, only selected apps, or all except selected apps, plus a
first-class "Exclude screen reader (VoiceOver) audio" toggle.

ScreenCaptureKit filters audio per application, so ScreenAudioCapture now
takes a ScreenAudioSelection and builds the matching SCContentFilter
(including:/excludingApplications:). New ScreenSharePickerSheet lists
running apps from SCShareableContent. iOS left untouched -- ReplayKit only
delivers the mixed system stream, so per-app filtering is impossible there.

clients/apple/macOS/VoiceCatMac/Audio/ScreenAudioCapture.swift

@@ -1,6 +1,25 @@
 import AVFoundation
 import ScreenCaptureKit
 
+// Which apps' audio the SCREEN_AUDIO stream captures. ScreenCaptureKit filters audio at the
+// *application* level (not per-window), so the selection is expressed as bundle IDs. The
+// picker UI (ScreenSharePickerSheet) produces a `ScreenAudioSelection`; `start()` turns it
+// into the matching `SCContentFilter`.
+enum ScreenAudioScope: Equatable {
+    case entireDesktop          // whole display — the original behaviour
+    case onlyApps([String])     // capture only these bundle IDs
+    case allExcept([String])    // capture everything except these bundle IDs
+}
+
+struct ScreenAudioSelection: Equatable {
+    var scope: ScreenAudioScope = .entireDesktop
+    /// Drop the macOS screen-reader (VoiceOver) speech from the shared mix. Meaningful for
+    /// `.entireDesktop`/`.allExcept`; for `.onlyApps` the screen reader is already excluded.
+    var excludeScreenReader: Bool = false
+
+    static let `default` = ScreenAudioSelection()
+}
+
 // ScreenAudioCapture — macOS system/desktop audio capture for the SCREEN_AUDIO stream.
 //
 // The macOS analog of the Windows WASAPI loopback path (docs/voice.md §9). ScreenCaptureKit
@@ -26,16 +45,27 @@ final class ScreenAudioCapture: NSObject, SCStreamOutput, SCStreamDelegate {
 
     private let onPcm: PcmHandler
     private let channels: Int  // 1 (mono) or 2 (stereo interleaved), matches the stream's mode
+    private let selection: ScreenAudioSelection
     private let sampleQueue = DispatchQueue(label: "cat.voice.screenaudio.samples")
     private var stream: SCStream?
 
+    /// Bundle IDs whose audio carries the macOS screen-reader speech. VoiceOver itself plus the
+    /// speech-synthesis daemon that actually renders the spoken audio — the speech is usually
+    /// emitted by the daemon, not the VoiceOver app, so we exclude whichever are running.
+    static let screenReaderBundleIDs: Set<String> = [
+        "com.apple.VoiceOver",
+        "com.apple.VoiceOver4",
+        "com.apple.speech.speechsynthesisd",
+    ]
+
     /// Interleaved int16 carry-over between callbacks (ScreenCaptureKit buffers don't align to
     /// 20 ms), drained in whole `frameSamplesPerChannel * channels` chunks. Only touched on
     /// `sampleQueue`.
     private var pending: [Int16] = []
 
-    init(channels: UInt32, onPcm: @escaping PcmHandler) {
+    init(channels: UInt32, selection: ScreenAudioSelection, onPcm: @escaping PcmHandler) {
         self.channels = max(1, min(2, Int(channels)))
+        self.selection = selection
         self.onPcm = onPcm
         super.init()
     }
@@ -46,7 +76,8 @@ final class ScreenAudioCapture: NSObject, SCStreamOutput, SCStreamDelegate {
         let content = try await SCShareableContent.current
         guard let display = content.displays.first else { throw CaptureError.noDisplay }
 
-        let filter = SCContentFilter(display: display, excludingWindows: [])
+        let filter = Self.makeFilter(selection: selection, display: display,
+                                     apps: content.applications)
 
         let config = SCStreamConfiguration()
         config.capturesAudio = true
@@ -65,6 +96,39 @@ final class ScreenAudioCapture: NSObject, SCStreamOutput, SCStreamDelegate {
         self.stream = stream
     }
 
+    /// Turn a `ScreenAudioSelection` into an `SCContentFilter` against the running apps.
+    /// ScreenCaptureKit filters audio per application, so we map bundle IDs → SCRunningApplication.
+    private static func makeFilter(selection: ScreenAudioSelection, display: SCDisplay,
+                                   apps: [SCRunningApplication]) -> SCContentFilter {
+        func appsMatching(_ ids: Set<String>) -> [SCRunningApplication] {
+            apps.filter { ids.contains($0.bundleIdentifier) }
+        }
+
+        switch selection.scope {
+        case .onlyApps(let bundleIDs):
+            // Include-only already excludes everything else (the screen reader included), so the
+            // excludeScreenReader flag is moot in this mode.
+            return SCContentFilter(display: display,
+                                   including: appsMatching(Set(bundleIDs)),
+                                   exceptingWindows: [])
+
+        case .allExcept(let bundleIDs):
+            var ids = Set(bundleIDs)
+            if selection.excludeScreenReader { ids.formUnion(screenReaderBundleIDs) }
+            return SCContentFilter(display: display,
+                                   excludingApplications: appsMatching(ids),
+                                   exceptingWindows: [])
+
+        case .entireDesktop:
+            if selection.excludeScreenReader {
+                return SCContentFilter(display: display,
+                                       excludingApplications: appsMatching(screenReaderBundleIDs),
+                                       exceptingWindows: [])
+            }
+            return SCContentFilter(display: display, excludingWindows: [])
+        }
+    }
+
     /// Stop capture and release the stream. Safe to call multiple times.
     func stop() {
         guard let stream else { return }