我有下面的JSON结构，我想用Logstash解析：

{
  "messages" : [ {
    "remoteReferenceId" : "133883",
    "sender" : {
      "name" : "User1"
    }
  }, {
    "remoteReferenceId" : "133894",
    "sender" : {
      "name" : "User2"
    }
  } ],
  "timestamp" : "2021-03-20T15:36:01.868+02:00"
}

我的配置如下：

input {
  file {
    mode => "read"
    path => "/path-to-data/*.json"
    file_completed_action => "delete"
    start_position => "beginning"
    sincedb_path => "/dev/null"
    type => "json"
    codec => multiline {
        pattern => "remoteReferenceId"
        negate => true
        what => previous
        auto_flush_interval => 1
        multiline_tag => ""
     }
  }
}
filter{
    json{
        source => "messages"
        target => "parsedJson"
    }
    mutate {
      add_field => {
        "remoteReferenceId" => "%{[remoteReferenceId]}"
          }
        }
}
output {
    elasticsearch{
        hosts => ["http://localhost:9200/"]
        index => "index_messages"
    }
    stdout { codec => rubydebug }
}

JSON文件被导入到Elasticsearch中。我的问题是，我想添加一个新字段“remoteReferenceId”，其中包含节点的值（例如上面的值“133883”），但在Elastic中，我得到的是“%{[remoteReferenceId]}”，而不是真实的的值。
有什么问题吗？非常感谢！