fix(android): remove mic and screen foreground services

2026-03-12 07:20:45 +00:00 · 2026-03-08 15:07:56 +05:30
parent 1230cefe25
commit 46145fde19
22 changed files with 30 additions and 371 deletions
--- a/apps/android/app/src/main/AndroidManifest.xml
+++ b/apps/android/app/src/main/AndroidManifest.xml
@@ -3,8 +3,6 @@
    <uses-permission android:name="android.permission.ACCESS_NETWORK_STATE" />
    <uses-permission android:name="android.permission.FOREGROUND_SERVICE" />
    <uses-permission android:name="android.permission.FOREGROUND_SERVICE_DATA_SYNC" />
-    <uses-permission android:name="android.permission.FOREGROUND_SERVICE_MICROPHONE" />
-    <uses-permission android:name="android.permission.FOREGROUND_SERVICE_MEDIA_PROJECTION" />
    <uses-permission android:name="android.permission.POST_NOTIFICATIONS" />
    <uses-permission
        android:name="android.permission.NEARBY_WIFI_DEVICES"
@@ -45,7 +43,7 @@
        <service
            android:name=".NodeForegroundService"
            android:exported="false"
-            android:foregroundServiceType="dataSync|microphone|mediaProjection" />
+            android:foregroundServiceType="dataSync" />
        <service
            android:name=".node.DeviceNotificationListenerService"
            android:label="@string/app_name"
--- a/apps/android/app/src/main/java/ai/openclaw/app/MainActivity.kt
+++ b/apps/android/app/src/main/java/ai/openclaw/app/MainActivity.kt
@@ -18,18 +18,14 @@ import kotlinx.coroutines.launch
 class MainActivity : ComponentActivity() {
  private val viewModel: MainViewModel by viewModels()
  private lateinit var permissionRequester: PermissionRequester
-  private lateinit var screenCaptureRequester: ScreenCaptureRequester

  override fun onCreate(savedInstanceState: Bundle?) {
    super.onCreate(savedInstanceState)
    WindowCompat.setDecorFitsSystemWindows(window, false)
    permissionRequester = PermissionRequester(this)
-    screenCaptureRequester = ScreenCaptureRequester(this)
    viewModel.camera.attachLifecycleOwner(this)
    viewModel.camera.attachPermissionRequester(permissionRequester)
    viewModel.sms.attachPermissionRequester(permissionRequester)
-    viewModel.screenRecorder.attachScreenCaptureRequester(screenCaptureRequester)
-    viewModel.screenRecorder.attachPermissionRequester(permissionRequester)

    lifecycleScope.launch {
      repeatOnLifecycle(Lifecycle.State.STARTED) {
--- a/apps/android/app/src/main/java/ai/openclaw/app/MainViewModel.kt
+++ b/apps/android/app/src/main/java/ai/openclaw/app/MainViewModel.kt
@@ -6,7 +6,6 @@ import ai.openclaw.app.gateway.GatewayEndpoint
 import ai.openclaw.app.chat.OutgoingAttachment
 import ai.openclaw.app.node.CameraCaptureManager
 import ai.openclaw.app.node.CanvasController
-import ai.openclaw.app.node.ScreenRecordManager
 import ai.openclaw.app.node.SmsManager
 import ai.openclaw.app.voice.VoiceConversationEntry
 import kotlinx.coroutines.flow.StateFlow
@@ -20,7 +19,6 @@ class MainViewModel(app: Application) : AndroidViewModel(app) {
  val canvasRehydratePending: StateFlow<Boolean> = runtime.canvasRehydratePending
  val canvasRehydrateErrorText: StateFlow<String?> = runtime.canvasRehydrateErrorText
  val camera: CameraCaptureManager = runtime.camera
-  val screenRecorder: ScreenRecordManager = runtime.screenRecorder
  val sms: SmsManager = runtime.sms

  val gateways: StateFlow<List<GatewayEndpoint>> = runtime.gateways
@@ -38,7 +36,6 @@ class MainViewModel(app: Application) : AndroidViewModel(app) {

  val cameraHud: StateFlow<CameraHudState?> = runtime.cameraHud
  val cameraFlashToken: StateFlow<Long> = runtime.cameraFlashToken
-  val screenRecordActive: StateFlow<Boolean> = runtime.screenRecordActive

  val instanceId: StateFlow<String> = runtime.instanceId
  val displayName: StateFlow<String> = runtime.displayName
--- a/apps/android/app/src/main/java/ai/openclaw/app/NodeForegroundService.kt
+++ b/apps/android/app/src/main/java/ai/openclaw/app/NodeForegroundService.kt
@@ -5,13 +5,10 @@ import android.app.NotificationChannel
 import android.app.NotificationManager
 import android.app.Service
 import android.app.PendingIntent
-import android.Manifest
 import android.content.Context
 import android.content.Intent
-import android.content.pm.PackageManager
 import android.content.pm.ServiceInfo
 import androidx.core.app.NotificationCompat
-import androidx.core.content.ContextCompat
 import kotlinx.coroutines.CoroutineScope
 import kotlinx.coroutines.Dispatchers
 import kotlinx.coroutines.Job
@@ -23,14 +20,13 @@ import kotlinx.coroutines.launch
 class NodeForegroundService : Service() {
  private val scope: CoroutineScope = CoroutineScope(SupervisorJob() + Dispatchers.Main)
  private var notificationJob: Job? = null
-  private var lastRequiresMic = false
  private var didStartForeground = false

  override fun onCreate() {
    super.onCreate()
    ensureChannel()
    val initial = buildNotification(title = "OpenClaw Node", text = "Starting…")
-    startForegroundWithTypes(notification = initial, requiresMic = false)
+    startForegroundWithTypes(notification = initial)

    val runtime = (application as NodeApp).runtime
    notificationJob =
@@ -53,11 +49,8 @@ class NodeForegroundService : Service() {
            }
          val text = (server?.let { "$status · $it" } ?: status) + micSuffix

-          val requiresMic =
-            micEnabled && hasRecordAudioPermission()
          startForegroundWithTypes(
            notification = buildNotification(title = title, text = text),
-            requiresMic = requiresMic,
          )
        }
      }
@@ -135,30 +128,15 @@ class NodeForegroundService : Service() {
    mgr.notify(NOTIFICATION_ID, notification)
  }

-  private fun startForegroundWithTypes(notification: Notification, requiresMic: Boolean) {
-    if (didStartForeground && requiresMic == lastRequiresMic) {
+  private fun startForegroundWithTypes(notification: Notification) {
+    if (didStartForeground) {
      updateNotification(notification)
      return
    }
-
-    lastRequiresMic = requiresMic
-    val types =
-      if (requiresMic) {
-        ServiceInfo.FOREGROUND_SERVICE_TYPE_DATA_SYNC or ServiceInfo.FOREGROUND_SERVICE_TYPE_MICROPHONE
-      } else {
-        ServiceInfo.FOREGROUND_SERVICE_TYPE_DATA_SYNC
-      }
-    startForeground(NOTIFICATION_ID, notification, types)
+    startForeground(NOTIFICATION_ID, notification, ServiceInfo.FOREGROUND_SERVICE_TYPE_DATA_SYNC)
    didStartForeground = true
  }

-  private fun hasRecordAudioPermission(): Boolean {
-    return (
-      ContextCompat.checkSelfPermission(this, Manifest.permission.RECORD_AUDIO) ==
-        PackageManager.PERMISSION_GRANTED
-      )
-  }
-
  companion object {
    private const val CHANNEL_ID = "connection"
    private const val NOTIFICATION_ID = 1
--- a/apps/android/app/src/main/java/ai/openclaw/app/NodeRuntime.kt
+++ b/apps/android/app/src/main/java/ai/openclaw/app/NodeRuntime.kt
@@ -50,7 +50,6 @@ class NodeRuntime(context: Context) {
  val canvas = CanvasController()
  val camera = CameraCaptureManager(appContext)
  val location = LocationCaptureManager(appContext)
-  val screenRecorder = ScreenRecordManager(appContext)
  val sms = SmsManager(appContext)
  private val json = Json { ignoreUnknownKeys = true }

@@ -113,12 +112,6 @@ class NodeRuntime(context: Context) {
    appContext = appContext,
  )

-  private val screenHandler: ScreenHandler = ScreenHandler(
-    screenRecorder = screenRecorder,
-    setScreenRecordActive = { _screenRecordActive.value = it },
-    invokeErrorFromThrowable = { invokeErrorFromThrowable(it) },
-  )
-
  private val smsHandlerImpl: SmsHandler = SmsHandler(
    sms = sms,
  )
@@ -153,7 +146,6 @@ class NodeRuntime(context: Context) {
    contactsHandler = contactsHandler,
    calendarHandler = calendarHandler,
    motionHandler = motionHandler,
-    screenHandler = screenHandler,
    smsHandler = smsHandlerImpl,
    a2uiHandler = a2uiHandler,
    debugHandler = debugHandler,
@@ -199,9 +191,6 @@ class NodeRuntime(context: Context) {
  private val _cameraFlashToken = MutableStateFlow(0L)
  val cameraFlashToken: StateFlow<Long> = _cameraFlashToken.asStateFlow()

-  private val _screenRecordActive = MutableStateFlow(false)
-  val screenRecordActive: StateFlow<Boolean> = _screenRecordActive.asStateFlow()
-
  private val _canvasA2uiHydrated = MutableStateFlow(false)
  val canvasA2uiHydrated: StateFlow<Boolean> = _canvasA2uiHydrated.asStateFlow()
  private val _canvasRehydratePending = MutableStateFlow(false)
@@ -616,6 +605,9 @@ class NodeRuntime(context: Context) {

  fun setForeground(value: Boolean) {
    _isForeground.value = value
+    if (!value) {
+      stopActiveVoiceSession()
+    }
  }

  fun setDisplayName(value: String) {
@@ -660,11 +652,7 @@ class NodeRuntime(context: Context) {

  fun setVoiceScreenActive(active: Boolean) {
    if (!active) {
-      // User left voice screen — stop mic and TTS
-      talkMode.ttsOnAllResponses = false
-      talkMode.stopTts()
-      micCapture.setMicEnabled(false)
-      prefs.setTalkEnabled(false)
+      stopActiveVoiceSession()
    }
    // Don't re-enable on active=true; mic toggle drives that
  }
@@ -693,6 +681,14 @@ class NodeRuntime(context: Context) {
    talkMode.setPlaybackEnabled(value)
  }

+  private fun stopActiveVoiceSession() {
+    talkMode.ttsOnAllResponses = false
+    talkMode.stopTts()
+    micCapture.setMicEnabled(false)
+    prefs.setTalkEnabled(false)
+    externalAudioCaptureActive.value = false
+  }
+
  fun refreshGatewayConnection() {
    val endpoint =
      connectedEndpoint ?: run {
--- a/apps/android/app/src/main/java/ai/openclaw/app/ScreenCaptureRequester.kt
+++ b/apps/android/app/src/main/java/ai/openclaw/app/ScreenCaptureRequester.kt
@@ -1,65 +0,0 @@
-package ai.openclaw.app
-
-import android.app.Activity
-import android.content.Context
-import android.content.Intent
-import android.media.projection.MediaProjectionManager
-import androidx.activity.ComponentActivity
-import androidx.activity.result.ActivityResultLauncher
-import androidx.activity.result.contract.ActivityResultContracts
-import androidx.appcompat.app.AlertDialog
-import kotlinx.coroutines.CompletableDeferred
-import kotlinx.coroutines.Dispatchers
-import kotlinx.coroutines.sync.Mutex
-import kotlinx.coroutines.sync.withLock
-import kotlinx.coroutines.withContext
-import kotlinx.coroutines.withTimeout
-import kotlinx.coroutines.suspendCancellableCoroutine
-import kotlin.coroutines.resume
-
-class ScreenCaptureRequester(private val activity: ComponentActivity) {
-  data class CaptureResult(val resultCode: Int, val data: Intent)
-
-  private val mutex = Mutex()
-  private var pending: CompletableDeferred<CaptureResult?>? = null
-
-  private val launcher: ActivityResultLauncher<Intent> =
-    activity.registerForActivityResult(ActivityResultContracts.StartActivityForResult()) { result ->
-      val p = pending
-      pending = null
-      val data = result.data
-      if (result.resultCode == Activity.RESULT_OK && data != null) {
-        p?.complete(CaptureResult(result.resultCode, data))
-      } else {
-        p?.complete(null)
-      }
-    }
-
-  suspend fun requestCapture(timeoutMs: Long = 20_000): CaptureResult? =
-    mutex.withLock {
-      val proceed = showRationaleDialog()
-      if (!proceed) return null
-
-      val mgr = activity.getSystemService(Context.MEDIA_PROJECTION_SERVICE) as MediaProjectionManager
-      val intent = mgr.createScreenCaptureIntent()
-
-      val deferred = CompletableDeferred<CaptureResult?>()
-      pending = deferred
-      withContext(Dispatchers.Main) { launcher.launch(intent) }
-
-      withContext(Dispatchers.Default) { withTimeout(timeoutMs) { deferred.await() } }
-    }
-
-  private suspend fun showRationaleDialog(): Boolean =
-    withContext(Dispatchers.Main) {
-      suspendCancellableCoroutine { cont ->
-        AlertDialog.Builder(activity)
-          .setTitle("Screen recording required")
-          .setMessage("OpenClaw needs to record the screen for this command.")
-          .setPositiveButton("Continue") { _, _ -> cont.resume(true) }
-          .setNegativeButton("Not now") { _, _ -> cont.resume(false) }
-          .setOnCancelListener { cont.resume(false) }
-          .show()
-      }
-    }
-}
--- a/apps/android/app/src/main/java/ai/openclaw/app/node/DeviceHandler.kt
+++ b/apps/android/app/src/main/java/ai/openclaw/app/node/DeviceHandler.kt
@@ -219,14 +219,6 @@ class DeviceHandler(
              promptableWhenDenied = true,
            ),
          )
-          // Screen capture on Android is interactive per-capture consent, not a sticky app permission.
-          put(
-            "screenCapture",
-            permissionStateJson(
-              granted = false,
-              promptableWhenDenied = true,
-            ),
-          )
        },
      )
    }.toString()
--- a/apps/android/app/src/main/java/ai/openclaw/app/node/InvokeCommandRegistry.kt
+++ b/apps/android/app/src/main/java/ai/openclaw/app/node/InvokeCommandRegistry.kt
@@ -11,7 +11,6 @@ import ai.openclaw.app.protocol.OpenClawLocationCommand
 import ai.openclaw.app.protocol.OpenClawMotionCommand
 import ai.openclaw.app.protocol.OpenClawNotificationsCommand
 import ai.openclaw.app.protocol.OpenClawPhotosCommand
-import ai.openclaw.app.protocol.OpenClawScreenCommand
 import ai.openclaw.app.protocol.OpenClawSmsCommand
 import ai.openclaw.app.protocol.OpenClawSystemCommand

@@ -59,7 +58,6 @@ object InvokeCommandRegistry {
  val capabilityManifest: List<NodeCapabilitySpec> =
    listOf(
      NodeCapabilitySpec(name = OpenClawCapability.Canvas.rawValue),
-      NodeCapabilitySpec(name = OpenClawCapability.Screen.rawValue),
      NodeCapabilitySpec(name = OpenClawCapability.Device.rawValue),
      NodeCapabilitySpec(name = OpenClawCapability.Notifications.rawValue),
      NodeCapabilitySpec(name = OpenClawCapability.System.rawValue),
@@ -122,10 +120,6 @@ object InvokeCommandRegistry {
        name = OpenClawCanvasA2UICommand.Reset.rawValue,
        requiresForeground = true,
      ),
-      InvokeCommandSpec(
-        name = OpenClawScreenCommand.Record.rawValue,
-        requiresForeground = true,
-      ),
      InvokeCommandSpec(
        name = OpenClawSystemCommand.Notify.rawValue,
      ),
--- a/apps/android/app/src/main/java/ai/openclaw/app/node/InvokeDispatcher.kt
+++ b/apps/android/app/src/main/java/ai/openclaw/app/node/InvokeDispatcher.kt
@@ -10,7 +10,6 @@ import ai.openclaw.app.protocol.OpenClawDeviceCommand
 import ai.openclaw.app.protocol.OpenClawLocationCommand
 import ai.openclaw.app.protocol.OpenClawMotionCommand
 import ai.openclaw.app.protocol.OpenClawNotificationsCommand
-import ai.openclaw.app.protocol.OpenClawScreenCommand
 import ai.openclaw.app.protocol.OpenClawSmsCommand
 import ai.openclaw.app.protocol.OpenClawSystemCommand

@@ -25,7 +24,6 @@ class InvokeDispatcher(
  private val contactsHandler: ContactsHandler,
  private val calendarHandler: CalendarHandler,
  private val motionHandler: MotionHandler,
-  private val screenHandler: ScreenHandler,
  private val smsHandler: SmsHandler,
  private val a2uiHandler: A2UIHandler,
  private val debugHandler: DebugHandler,
@@ -160,9 +158,6 @@ class InvokeDispatcher(
      OpenClawMotionCommand.Activity.rawValue -> motionHandler.handleMotionActivity(paramsJson)
      OpenClawMotionCommand.Pedometer.rawValue -> motionHandler.handleMotionPedometer(paramsJson)

-      // Screen command
-      OpenClawScreenCommand.Record.rawValue -> screenHandler.handleScreenRecord(paramsJson)
-
      // SMS command
      OpenClawSmsCommand.Send.rawValue -> smsHandler.handleSmsSend(paramsJson)

--- a/apps/android/app/src/main/java/ai/openclaw/app/node/ScreenHandler.kt
+++ b/apps/android/app/src/main/java/ai/openclaw/app/node/ScreenHandler.kt
@@ -1,25 +0,0 @@
-package ai.openclaw.app.node
-
-import ai.openclaw.app.gateway.GatewaySession
-
-class ScreenHandler(
-  private val screenRecorder: ScreenRecordManager,
-  private val setScreenRecordActive: (Boolean) -> Unit,
-  private val invokeErrorFromThrowable: (Throwable) -> Pair<String, String>,
-) {
-  suspend fun handleScreenRecord(paramsJson: String?): GatewaySession.InvokeResult {
-    setScreenRecordActive(true)
-    try {
-      val res =
-        try {
-          screenRecorder.record(paramsJson)
-        } catch (err: Throwable) {
-          val (code, message) = invokeErrorFromThrowable(err)
-          return GatewaySession.InvokeResult.error(code = code, message = message)
-        }
-      return GatewaySession.InvokeResult.ok(res.payloadJson)
-    } finally {
-      setScreenRecordActive(false)
-    }
-  }
-}
--- a/apps/android/app/src/main/java/ai/openclaw/app/node/ScreenRecordManager.kt
+++ b/apps/android/app/src/main/java/ai/openclaw/app/node/ScreenRecordManager.kt
@@ -1,165 +0,0 @@
-package ai.openclaw.app.node
-
-import android.content.Context
-import android.hardware.display.DisplayManager
-import android.media.MediaRecorder
-import android.media.projection.MediaProjectionManager
-import android.os.Build
-import android.util.Base64
-import ai.openclaw.app.ScreenCaptureRequester
-import kotlinx.coroutines.Dispatchers
-import kotlinx.coroutines.delay
-import kotlinx.coroutines.withContext
-import kotlinx.serialization.json.JsonObject
-import java.io.File
-import kotlin.math.roundToInt
-
-class ScreenRecordManager(private val context: Context) {
-  data class Payload(val payloadJson: String)
-
-  @Volatile private var screenCaptureRequester: ScreenCaptureRequester? = null
-  @Volatile private var permissionRequester: ai.openclaw.app.PermissionRequester? = null
-
-  fun attachScreenCaptureRequester(requester: ScreenCaptureRequester) {
-    screenCaptureRequester = requester
-  }
-
-  fun attachPermissionRequester(requester: ai.openclaw.app.PermissionRequester) {
-    permissionRequester = requester
-  }
-
-  suspend fun record(paramsJson: String?): Payload =
-    withContext(Dispatchers.Default) {
-      val requester =
-        screenCaptureRequester
-          ?: throw IllegalStateException(
-            "SCREEN_PERMISSION_REQUIRED: grant Screen Recording permission",
-          )
-
-      val params = parseJsonParamsObject(paramsJson)
-      val durationMs = (parseDurationMs(params) ?: 10_000).coerceIn(250, 60_000)
-      val fps = (parseFps(params) ?: 10.0).coerceIn(1.0, 60.0)
-      val fpsInt = fps.roundToInt().coerceIn(1, 60)
-      val screenIndex = parseScreenIndex(params)
-      val includeAudio = parseIncludeAudio(params) ?: true
-      val format = parseString(params, key = "format")
-      if (format != null && format.lowercase() != "mp4") {
-        throw IllegalArgumentException("INVALID_REQUEST: screen format must be mp4")
-      }
-      if (screenIndex != null && screenIndex != 0) {
-        throw IllegalArgumentException("INVALID_REQUEST: screenIndex must be 0 on Android")
-      }
-
-      val capture = requester.requestCapture()
-        ?: throw IllegalStateException(
-          "SCREEN_PERMISSION_REQUIRED: grant Screen Recording permission",
-        )
-
-      val mgr =
-        context.getSystemService(Context.MEDIA_PROJECTION_SERVICE) as MediaProjectionManager
-      val projection = mgr.getMediaProjection(capture.resultCode, capture.data)
-        ?: throw IllegalStateException("UNAVAILABLE: screen capture unavailable")
-
-      val metrics = context.resources.displayMetrics
-      val width = metrics.widthPixels
-      val height = metrics.heightPixels
-      val densityDpi = metrics.densityDpi
-
-      val file = File.createTempFile("openclaw-screen-", ".mp4")
-      if (includeAudio) ensureMicPermission()
-
-      val recorder = createMediaRecorder()
-      var virtualDisplay: android.hardware.display.VirtualDisplay? = null
-      try {
-        if (includeAudio) {
-          recorder.setAudioSource(MediaRecorder.AudioSource.MIC)
-        }
-        recorder.setVideoSource(MediaRecorder.VideoSource.SURFACE)
-        recorder.setOutputFormat(MediaRecorder.OutputFormat.MPEG_4)
-        recorder.setVideoEncoder(MediaRecorder.VideoEncoder.H264)
-        if (includeAudio) {
-          recorder.setAudioEncoder(MediaRecorder.AudioEncoder.AAC)
-          recorder.setAudioChannels(1)
-          recorder.setAudioSamplingRate(44_100)
-          recorder.setAudioEncodingBitRate(96_000)
-        }
-        recorder.setVideoSize(width, height)
-        recorder.setVideoFrameRate(fpsInt)
-        recorder.setVideoEncodingBitRate(estimateBitrate(width, height, fpsInt))
-        recorder.setOutputFile(file.absolutePath)
-        recorder.prepare()
-
-        val surface = recorder.surface
-        virtualDisplay =
-          projection.createVirtualDisplay(
-            "openclaw-screen",
-            width,
-            height,
-            densityDpi,
-            DisplayManager.VIRTUAL_DISPLAY_FLAG_AUTO_MIRROR,
-            surface,
-            null,
-            null,
-          )
-
-        recorder.start()
-        delay(durationMs.toLong())
-      } finally {
-        try {
-          recorder.stop()
-        } catch (_: Throwable) {
-          // ignore
-        }
-        recorder.reset()
-        recorder.release()
-        virtualDisplay?.release()
-        projection.stop()
-      }
-
-      val bytes = withContext(Dispatchers.IO) { file.readBytes() }
-      file.delete()
-      val base64 = Base64.encodeToString(bytes, Base64.NO_WRAP)
-      Payload(
-        """{"format":"mp4","base64":"$base64","durationMs":$durationMs,"fps":$fpsInt,"screenIndex":0,"hasAudio":$includeAudio}""",
-      )
-    }
-
-  private fun createMediaRecorder(): MediaRecorder = MediaRecorder(context)
-
-  private suspend fun ensureMicPermission() {
-    val granted =
-      androidx.core.content.ContextCompat.checkSelfPermission(
-        context,
-        android.Manifest.permission.RECORD_AUDIO,
-      ) == android.content.pm.PackageManager.PERMISSION_GRANTED
-    if (granted) return
-
-    val requester =
-      permissionRequester
-        ?: throw IllegalStateException("MIC_PERMISSION_REQUIRED: grant Microphone permission")
-    val results = requester.requestIfMissing(listOf(android.Manifest.permission.RECORD_AUDIO))
-    if (results[android.Manifest.permission.RECORD_AUDIO] != true) {
-      throw IllegalStateException("MIC_PERMISSION_REQUIRED: grant Microphone permission")
-    }
-  }
-
-  private fun parseDurationMs(params: JsonObject?): Int? =
-    parseJsonInt(params, "durationMs")
-
-  private fun parseFps(params: JsonObject?): Double? =
-    parseJsonDouble(params, "fps")
-
-  private fun parseScreenIndex(params: JsonObject?): Int? =
-    parseJsonInt(params, "screenIndex")
-
-  private fun parseIncludeAudio(params: JsonObject?): Boolean? = parseJsonBooleanFlag(params, "includeAudio")
-
-  private fun parseString(params: JsonObject?, key: String): String? =
-    parseJsonString(params, key)
-
-  private fun estimateBitrate(width: Int, height: Int, fps: Int): Int {
-    val pixels = width.toLong() * height.toLong()
-    val raw = (pixels * fps.toLong() * 2L).toInt()
-    return raw.coerceIn(1_000_000, 12_000_000)
-  }
-}
--- a/apps/android/app/src/main/java/ai/openclaw/app/protocol/OpenClawProtocolConstants.kt
+++ b/apps/android/app/src/main/java/ai/openclaw/app/protocol/OpenClawProtocolConstants.kt
@@ -3,7 +3,6 @@ package ai.openclaw.app.protocol
 enum class OpenClawCapability(val rawValue: String) {
  Canvas("canvas"),
  Camera("camera"),
-  Screen("screen"),
  Sms("sms"),
  VoiceWake("voiceWake"),
  Location("location"),
@@ -51,15 +50,6 @@ enum class OpenClawCameraCommand(val rawValue: String) {
  }
 }

-enum class OpenClawScreenCommand(val rawValue: String) {
-  Record("screen.record"),
-  ;
-
-  companion object {
-    const val NamespacePrefix: String = "screen."
-  }
-}
-
 enum class OpenClawSmsCommand(val rawValue: String) {
  Send("sms.send"),
  ;
--- a/apps/android/app/src/main/java/ai/openclaw/app/ui/OnboardingFlow.kt
+++ b/apps/android/app/src/main/java/ai/openclaw/app/ui/OnboardingFlow.kt
@@ -1402,7 +1402,7 @@ private fun PermissionsStep(
    InlineDivider()
    PermissionToggleRow(
      title = "Microphone",
-      subtitle = "Voice tab transcription",
+      subtitle = "Foreground Voice tab transcription",
      checked = enableMicrophone,
      granted = isPermissionGranted(context, Manifest.permission.RECORD_AUDIO),
      onCheckedChange = onMicrophoneChange,
--- a/apps/android/app/src/main/java/ai/openclaw/app/ui/SettingsSheet.kt
+++ b/apps/android/app/src/main/java/ai/openclaw/app/ui/SettingsSheet.kt
@@ -402,9 +402,9 @@ fun SettingsSheet(viewModel: MainViewModel) {
          supportingContent = {
            Text(
              if (micPermissionGranted) {
-                "Granted. Use the Voice tab mic button to capture transcript."
+                "Granted. Use the Voice tab mic button to capture transcript while the app is open."
              } else {
-                "Required for Voice tab transcription."
+                "Required for foreground Voice tab transcription."
              },
              style = mobileCallout,
            )
@@ -431,7 +431,7 @@ fun SettingsSheet(viewModel: MainViewModel) {
      }
      item {
        Text(
-          "Voice wake and talk modes were removed. Voice now uses one mic on/off flow in the Voice tab.",
+          "Voice wake and talk modes were removed. Voice now uses one mic on/off flow in the Voice tab while the app is open.",
          style = mobileCallout,
          color = mobileTextSecondary,
        )
--- a/apps/android/app/src/test/java/ai/openclaw/app/node/DeviceHandlerTest.kt
+++ b/apps/android/app/src/test/java/ai/openclaw/app/node/DeviceHandlerTest.kt
@@ -94,7 +94,6 @@ class DeviceHandlerTest {
        "contacts",
        "calendar",
        "motion",
-        "screenCapture",
      )
    for (key in expected) {
      val state = permissions.getValue(key).jsonObject
--- a/apps/android/app/src/test/java/ai/openclaw/app/node/InvokeCommandRegistryTest.kt
+++ b/apps/android/app/src/test/java/ai/openclaw/app/node/InvokeCommandRegistryTest.kt
@@ -19,7 +19,6 @@ class InvokeCommandRegistryTest {
  private val coreCapabilities =
    setOf(
      OpenClawCapability.Canvas.rawValue,
-      OpenClawCapability.Screen.rawValue,
      OpenClawCapability.Device.rawValue,
      OpenClawCapability.Notifications.rawValue,
      OpenClawCapability.System.rawValue,
--- a/apps/android/app/src/test/java/ai/openclaw/app/protocol/OpenClawProtocolConstantsTest.kt
+++ b/apps/android/app/src/test/java/ai/openclaw/app/protocol/OpenClawProtocolConstantsTest.kt
@@ -24,7 +24,6 @@ class OpenClawProtocolConstantsTest {
  fun capabilitiesUseStableStrings() {
    assertEquals("canvas", OpenClawCapability.Canvas.rawValue)
    assertEquals("camera", OpenClawCapability.Camera.rawValue)
-    assertEquals("screen", OpenClawCapability.Screen.rawValue)
    assertEquals("voiceWake", OpenClawCapability.VoiceWake.rawValue)
    assertEquals("location", OpenClawCapability.Location.rawValue)
    assertEquals("sms", OpenClawCapability.Sms.rawValue)
@@ -44,11 +43,6 @@ class OpenClawProtocolConstantsTest {
    assertEquals("camera.clip", OpenClawCameraCommand.Clip.rawValue)
  }

-  @Test
-  fun screenCommandsUseStableStrings() {
-    assertEquals("screen.record", OpenClawScreenCommand.Record.rawValue)
-  }
-
  @Test
  fun notificationsCommandsUseStableStrings() {
    assertEquals("notifications.list", OpenClawNotificationsCommand.List.rawValue)
--- a/docs/concepts/features.md
+++ b/docs/concepts/features.md
@@ -45,7 +45,7 @@ title: "Features"
 - Optional voice note transcription hook
 - WebChat and macOS menu bar app
 - iOS node with pairing, Canvas, camera, screen recording, location, and voice features
- Android node with pairing, Connect tab, chat sessions, voice tab, Canvas/camera/screen, plus device, notifications, contacts/calendar, motion, photos, and SMS commands
+- Android node with pairing, Connect tab, chat sessions, voice tab, Canvas/camera, plus device, notifications, contacts/calendar, motion, photos, and SMS commands

 <Note>
 Legacy Claude, Codex, Gemini, and Opencode paths have been removed. Pi is the only
--- a/docs/index.md
+++ b/docs/index.md
@@ -89,7 +89,7 @@ The Gateway is the single source of truth for sessions, routing, and channel con
    Browser dashboard for chat, config, sessions, and nodes.
  </Card>
  <Card title="Mobile nodes" icon="smartphone">
-    Pair iOS and Android nodes for Canvas, camera/screen, and voice-enabled workflows.
+    Pair iOS and Android nodes for Canvas, camera, and voice-enabled workflows.
  </Card>
 </Columns>

@@ -164,7 +164,7 @@ Example:
    Channel-specific setup for WhatsApp, Telegram, Discord, and more.
  </Card>
  <Card title="Nodes" href="/nodes" icon="smartphone">
-    iOS and Android nodes with pairing, Canvas, camera/screen, and device actions.
+    iOS and Android nodes with pairing, Canvas, camera, and device actions.
  </Card>
  <Card title="Help" href="/help" icon="life-buoy">
    Common fixes and troubleshooting entry point.
--- a/docs/nodes/index.md
+++ b/docs/nodes/index.md
@@ -216,7 +216,7 @@ Notes:

 ## Screen recordings (nodes)

-Nodes expose `screen.record` (mp4). Example:
+Supported nodes expose `screen.record` (mp4). Example:

 ```bash
 openclaw nodes screen record --node <idOrNameOrIp> --duration 10s --fps 10
@@ -225,10 +225,9 @@ openclaw nodes screen record --node <idOrNameOrIp> --duration 10s --fps 10 --no-

 Notes:

- `screen.record` requires the node app to be foregrounded.
- Android will show the system screen-capture prompt before recording.
+- `screen.record` availability depends on node platform.
 - Screen recordings are clamped to `<= 60s`.
- `--no-audio` disables microphone capture (supported on iOS/Android; macOS uses system capture audio).
+- `--no-audio` disables microphone capture on supported platforms.
 - Use `--screen <index>` to select a display when multiple screens are available.

 ## Location (nodes)
--- a/docs/platforms/android.md
+++ b/docs/platforms/android.md
@@ -118,7 +118,7 @@ The Android Chat tab supports session selection (default `main`, plus other exis
 - Send: `chat.send`
 - Push updates (best-effort): `chat.subscribe` → `event:"chat"`

-### 7) Canvas + screen + camera
+### 7) Canvas + camera

 #### Gateway Canvas Host (recommended for web content)

@@ -151,13 +151,9 @@ Camera commands (foreground only; permission-gated):

 See [Camera node](/nodes/camera) for parameters and CLI helpers.

-Screen commands:
-
- `screen.record` (mp4; foreground only)
-
 ### 8) Voice + expanded Android command surface

- Voice: Android uses a single mic on/off flow in the Voice tab with transcript capture and TTS playback (ElevenLabs when configured, system TTS fallback).
+- Voice: Android uses a single mic on/off flow in the Voice tab with transcript capture and TTS playback (ElevenLabs when configured, system TTS fallback). Voice stops when the app leaves the foreground.
 - Voice wake/talk-mode toggles are currently removed from Android UX/runtime.
 - Additional Android command families (availability depends on device + permissions):
  - `device.status`, `device.info`, `device.permissions`, `device.health`
--- a/src/gateway/android-node.capabilities.live.test.ts
+++ b/src/gateway/android-node.capabilities.live.test.ts
@@ -12,7 +12,7 @@ import { resolveGatewayCredentialsFromConfig } from "./credentials.js";
 const LIVE = isTruthyEnvValue(process.env.LIVE) || isTruthyEnvValue(process.env.OPENCLAW_LIVE_TEST);
 const LIVE_ANDROID_NODE = isTruthyEnvValue(process.env.OPENCLAW_LIVE_ANDROID_NODE);
 const describeLive = LIVE && LIVE_ANDROID_NODE ? describe : describe.skip;
-const SKIPPED_INTERACTIVE_COMMANDS = new Set<string>(["screen.record"]);
+const SKIPPED_INTERACTIVE_COMMANDS = new Set<string>();

 type CommandOutcome = "success" | "error";

@@ -120,15 +120,6 @@ const COMMAND_PROFILES: Record<string, CommandProfile> = {
    timeoutMs: 30_000,
    outcome: "success",
  },
-  "screen.record": {
-    buildParams: () => ({ durationMs: 1500, fps: 8, includeAudio: false }),
-    timeoutMs: 60_000,
-    outcome: "success",
-    onSuccess: (payload) => {
-      const obj = assertObjectPayload("screen.record", payload);
-      expect(readString(obj.base64)).not.toBeNull();
-    },
-  },
  "camera.list": {
    buildParams: () => ({}),
    timeoutMs: 20_000,