public Task<bool> WaitForRowsAsync()

{
  await QueuedTask.Run(async () =>
  {
    //and assuming you have established a connection to a knowledge graph
    //...
    //Construct an openCypher query - return the first 10 entities (whatever
    //they are...)
    var query = "MATCH (n) RETURN n LIMIT 10";//default limit is 100 if not specified
                                              //other examples...
                                              //query = "MATCH (a:Person) RETURN [a.name, a.age] ORDER BY a.age DESC LIMIT 50";
                                              //query = "MATCH (b:Person) RETURN { Xperson: { Xname: b.name, Xage: b.age } } ORDER BY b.name DESC";
                                              //query = "MATCH p = (c:Person)-[:HasCar]-() RETURN p ORDER BY c.name DESC";

    //Create a query filter
    //Note: OutputSpatialReference is currently ignored
    var kg_qf = new KnowledgeGraphQueryFilter()
    {
      QueryText = query
    };
    //Optionally - u can choose to include provenance in the results
    //(_if_ the KG has provenance - otherwise the query will fail)
    bool includeProvenanceIfPresent = KnowledgeGraphSupportsProvenance(kg);

    if (includeProvenanceIfPresent)
    {
      //Only include if the KG has provenance
      kg_qf.ProvenanceBehavior =
          KnowledgeGraphProvenanceBehavior.Include;//default is exclude
    }
    //submit the query - returns a KnowledgeGraphCursor
    using (var kg_rc = kg.SubmitQuery(kg_qf))
    {
      //wait for rows to be returned from the server
      //note the "await"...
      while (await kg_rc.WaitForRowsAsync())
      {
        //Rows have been retrieved - process this "batch"...
        while (kg_rc.MoveNext())
        {
          //Get the current KnowledgeGraphRow
          using (var graph_row = kg_rc.Current)
          {
            //Graph row is an array, process all returned values...
            var val_count = (int)graph_row.GetCount();
            for (int i = 0; i < val_count; i++)
            {
              var retval = graph_row[i];
              //Process row value (note: recursive)
              //See "Process a KnowledgeGraphRow Value" snippet
              ProcessKnowledgeGraphRowValue(retval);
            }
          }
        }
      }//WaitForRowsAsync
    }//SubmitQuery
  });
}

{
  // Note: call within QueuedTask.Run()
  {
    //query filter can be null to search and retrieve all rows
    //true means recycling cursor
    using (var rc = streamLayer.SearchAndSubscribe(qfilter, true))
    {
      //waiting for new features to be streamed
      //default is no cancellation
      while (rc.WaitForRowsAsync().Result)
      {
        while (rc.MoveNext())
        {
          using (var row = rc.Current)
          {
            //determine the origin of the row event
            switch (row.GetRowSource())
            {
              case RealtimeRowSource.PreExisting:
                //pre-existing row at the time of subscribe
                continue;
              case RealtimeRowSource.EventInsert:
                //row was inserted after subscribe
                continue;
              case RealtimeRowSource.EventDelete:
                //row was deleted after subscribe
                continue;
            }
          }
        }
      }
    }//row cursor is disposed. row cursor is unsubscribed

    //....or....
    //Use the feature class instead of the layer
    using var rfc = streamLayer.GetFeatureClass();
    //non-recycling cursor - 2nd param "false"
    using (RealtimeCursor rc = rfc.SearchAndSubscribe(qfilter, false))
    {
      //waiting for new features to be streamed
      //default is no cancellation
      while (rc.WaitForRowsAsync().Result)
      {
        //etc
      }
    }
  }
}

{
  // Note: call within QueuedTask.Run()
  {
    //Recycling cursor - 2nd param "true"
    //or streamLayer.Subscribe(qfilter, true) to just subscribe
    using (var rc = streamLayer.SearchAndSubscribe(qfilter, true))
    {
      //auto-cancel after 20 seconds
      var cancel = new CancellationTokenSource(new TimeSpan(0, 0, 20));
      //catch TaskCanceledException
      try
      {
        while (rc.WaitForRowsAsync(cancel.Token).Result)
        {
          //check for row events
          while (rc.MoveNext())
          {
            using var row = rc.Current;
            //etc
          }
        }
      }
      catch (TaskCanceledException)
      {
        //Handle cancellation as needed
      }
      cancel.Dispose();
    }
  }
}

{
  //Note we can use System Task with the Realtime feature class
  //for subscribe

  // Note: call within QueuedTask.Run()
  {
    using RealtimeFeatureClass rfc = streamLayer.GetFeatureClass();
    //non-recycling cursor - 2nd param "false"
    using RealtimeCursor rc = rfc.SearchAndSubscribe(qfilter, false);

    //waiting for new features to be streamed
    //default is no cancellation - use await in async method
    while (rc.WaitForRowsAsync().Result)
    {
      //pre-existing rows will be retrieved that were searched
      while (rc.MoveNext())
      {
        using RealtimeRow row = rc.Current;
        var row_source = row.GetRowSource();
        switch (row_source)
        {
          case RealtimeRowSource.EventDelete:
            //TODO - handle deletes
            break;
          case RealtimeRowSource.EventInsert:
            //TODO handle inserts
            break;
          case RealtimeRowSource.PreExisting:
            //TODO handle pre-existing rows
            break;
        }
      }
    }
    //row cursor is disposed. row cursor is unsubscribed
  }
}