feat: add Maestro replay compatibility

ios-runner/AgentDeviceRunner/AgentDeviceRunnerUITests/RunnerTests+CommandExecution.swift

@@ -252,7 +252,12 @@ extension RunnerTests {
       )
     case .tap:
       if let selectorKey = command.selectorKey, let selectorValue = command.selectorValue {
-        let match = findElement(app: activeApp, selectorKey: selectorKey, selectorValue: selectorValue)
+        let match = findElement(
+          app: activeApp,
+          selectorKey: selectorKey,
+          selectorValue: selectorValue,
+          allowNonHittableFallback: command.allowNonHittableSelectorTap == true
+        )
         if match.isAmbiguous {
           return Response(ok: false, error: ErrorPayload(code: "AMBIGUOUS_MATCH", message: "selector matched multiple elements"))
         }
@@ -264,16 +269,24 @@ extension RunnerTests {
           var outcome = RunnerInteractionOutcome.performed
           let timing = measureGesture {
             withTemporaryScrollIdleTimeoutIfSupported(activeApp) {
-              outcome = activateElement(app: activeApp, element: element, action: "tap by selector")
+              if match.usedNonHittableFallback {
+                // Maestro compatibility: RN E2E backdoor controls can be 1x1 and
+                // reported non-hittable by XCTest, while Maestro still taps their
+                // resolved bounds. Keep this behind the explicit replay-only flag.
+                outcome = tapAt(app: activeApp, x: frame.midX, y: frame.midY)
+              } else {
+                outcome = activateElement(app: activeApp, element: element, action: "tap by selector")
+              }
             }
           }
           if let response = unsupportedResponse(for: outcome) {
             return response
           }
+          waitForTextEntryReadinessAfterTap(app: activeApp, element: element)
           return Response(
             ok: true,
             data: DataPayload(
-              message: "tapped",
+              message: match.usedNonHittableFallback ? "tapped via non-hittable coordinate fallback" : "tapped",
               gestureStartUptimeMs: timing.gestureStartUptimeMs,
               gestureEndUptimeMs: timing.gestureEndUptimeMs,
               x: touchFrame?.x,

ios-runner/AgentDeviceRunner/AgentDeviceRunnerUITests/RunnerTests+Interaction.swift

@@ -27,6 +27,7 @@ extension RunnerTests {
   struct SelectorElementMatch {
     let element: XCUIElement?
     let isAmbiguous: Bool
+    let usedNonHittableFallback: Bool
   }
 
   enum TextTypingRepairMode {
@@ -177,10 +178,15 @@ extension RunnerTests {
     return element.exists ? element : nil
   }
 
-  func findElement(app: XCUIApplication, selectorKey: String, selectorValue: String) -> SelectorElementMatch {
+  func findElement(
+    app: XCUIApplication,
+    selectorKey: String,
+    selectorValue: String,
+    allowNonHittableFallback: Bool = false
+  ) -> SelectorElementMatch {
     let value = selectorValue.trimmingCharacters(in: .whitespacesAndNewlines)
     guard !value.isEmpty else {
-      return SelectorElementMatch(element: nil, isAmbiguous: false)
+      return SelectorElementMatch(element: nil, isAmbiguous: false, usedNonHittableFallback: false)
     }
     let predicate: NSPredicate
     switch selectorKey {
@@ -193,21 +199,47 @@ extension RunnerTests {
     case "text":
       predicate = NSPredicate(format: "label ==[c] %@ OR identifier ==[c] %@ OR value ==[c] %@", value, value, value)
     default:
-      return SelectorElementMatch(element: nil, isAmbiguous: false)
+      return SelectorElementMatch(element: nil, isAmbiguous: false, usedNonHittableFallback: false)
     }
 
     var matchedElement: XCUIElement?
+    var nonHittableElement: XCUIElement?
     let matches = app.descendants(matching: .any).matching(predicate).allElementsBoundByIndex
     for element in matches where element.exists {
-      guard element.isHittable else {
+      if !element.isHittable {
+        if allowNonHittableFallback && hasTappableFrame(app: app, element: element) {
+          guard nonHittableElement == nil else {
+            return SelectorElementMatch(element: nil, isAmbiguous: true, usedNonHittableFallback: false)
+          }
+          nonHittableElement = element
+        }
         continue
       }
       guard matchedElement == nil else {
-        return SelectorElementMatch(element: nil, isAmbiguous: true)
+        return SelectorElementMatch(element: nil, isAmbiguous: true, usedNonHittableFallback: false)
       }
       matchedElement = element
     }
-    return SelectorElementMatch(element: matchedElement, isAmbiguous: false)
+    if let matchedElement {
+      return SelectorElementMatch(element: matchedElement, isAmbiguous: false, usedNonHittableFallback: false)
+    }
+    return SelectorElementMatch(
+      element: nonHittableElement,
+      isAmbiguous: false,
+      usedNonHittableFallback: nonHittableElement != nil
+    )
+  }
+
+  private func hasTappableFrame(app: XCUIApplication, element: XCUIElement) -> Bool {
+    let frame = element.frame
+    if frame.isEmpty {
+      return false
+    }
+    let appFrame = app.frame
+    if appFrame.isEmpty {
+      return true
+    }
+    return appFrame.contains(CGPoint(x: frame.midX, y: frame.midY))
   }
 
   func queryElement(app: XCUIApplication, selectorKey: String, selectorValue: String) -> Response {
@@ -780,6 +812,35 @@ extension RunnerTests {
 #endif
   }
 
+  func waitForTextEntryReadinessAfterTap(app: XCUIApplication, element: XCUIElement) {
+#if os(iOS)
+    switch element.elementType {
+    case .textField, .secureTextField, .searchField, .textView:
+      if waitForFocusedTextInput(app: app, timeout: TextEntryTiming.readinessTimeout) != nil {
+        return
+      }
+      let frame = element.frame
+      if !frame.isEmpty {
+        _ = tapAt(app: app, x: frame.midX, y: frame.midY)
+        _ = waitForFocusedTextInput(app: app, timeout: TextEntryTiming.readinessTimeout)
+      }
+    default:
+      return
+    }
+#endif
+  }
+
+  private func waitForFocusedTextInput(app: XCUIApplication, timeout: TimeInterval) -> XCUIElement? {
+    let deadline = Date().addingTimeInterval(timeout)
+    while Date() < deadline {
+      if let focused = focusedTextInput(app: app) {
+        return focused
+      }
+      sleepFor(TextEntryTiming.pollInterval)
+    }
+    return focusedTextInput(app: app)
+  }
+
   private func textEntryRefreshPoint(for element: XCUIElement?) -> CGPoint? {
     guard let element else {
       return nil

ios-runner/AgentDeviceRunner/AgentDeviceRunnerUITests/RunnerTests+Models.swift

@@ -37,6 +37,7 @@ struct Command: Codable {
   let text: String?
   let selectorKey: String?
   let selectorValue: String?
+  let allowNonHittableSelectorTap: Bool?
   let delayMs: Int?
   let textEntryMode: String?
   let clearFirst: Bool?

src/compat/__tests__/replay-input.test.ts

@@ -32,7 +32,7 @@ test('parseReplayInput routes compat replay scripts through the selected parser'
     parsed.actions.map((action) => [action.command, action.positionals]),
     [
       ['open', ['com.callstack.agentdevicelab']],
-      ['click', ['id="submit-order"']],
+      ['__maestroTapOn', ['id="submit-order"']],
     ],
   );
 });
@@ -60,7 +60,7 @@ env:
     parsed.actions.map((action) => [action.command, action.positionals]),
     [
       ['open', ['cli-app']],
-      ['click', ['id="shell-button"']],
+      ['__maestroTapOn', ['id="shell-button"']],
     ],
   );
 });