Attacks in RFID systems are known to be expressed in output data. As these systems become more widespread, research into detecting attacks in the output data which is produced when RFID components interact will become more important. However, the problem limiting practical research developments in this area has been the availability of output data containing attacks. As the output data is contextualized by the application environment which produced it, detecting attacks requires knowledge of the environment to construct relevant plausibility checks. To overcome this problem we introduce a software based RFID simulation platform called Tyrell. The end-user instantiates the values of a real system and attacks into the simulation platform. The interaction between the simulated components leads to the production of output data containing attacks. The improvement is that output data can now be produced containing attacks in a repeatable way.