Dry Run

A safety mechanism that simulates execution and outputs the expected results before an operation is actually performed, allowing users to review the list of actions an agent intends to take. It is commonly used for human confirmation before dangerous or irreversible operations.