EAGLET is a tool to evaluate the gold standards used for NER and EL tasks based on a set of uniform annotation rules. It provides a very generic, adaptable set of rules derived from existing benchmarks allowing a check for these gold standards in a semi-automatic way. The systematic classification of errors above allows for the creation of a framework, which can detect and correct a large portion of anomalies which are in violation of the rules.