Systems and methods are provided for detecting an actual or potential cascade of data rules within a system that processes data, or a cascade of data operations caused by the rules. A cascade involves multiple rules such that an action or actions performed by one rule upon a set of data (e.g., when specified conditions are satisfied) trigger automatic execution of another rule that performs one or more additional actions. Depending on the length and complexity of a cascade, which may be created accidentally, significant amounts of computing resources may be wasted, particularly if the cascade creates a loop that, if unresolved, would cause the cascade to execute repeatedly. One or more temporary data structures are employed to track the effects of a collection of rules (e.g., changes to data item values) to determine whether and when one rule's action(s) trigger another rule.