What feature in dbt allows for the reuse of common transformations across models?

The feature in dbt that allows for the reuse of common transformations across models is macros. Macros are reusable pieces of SQL code that can be defined in your dbt project, enabling you to encapsulate complex logic or common transformations. Once a macro is created, it can be invoked in multiple models or other macros, which promotes DRY (Don't Repeat Yourself) principles and streamlines maintenance. This makes it easier for data teams to implement consistent transformations without duplicating code, leading to cleaner and more maintainable projects.

In contrast, models represent the tables and views in your data warehouse and do not inherently support code reuse like macros do. Migrations are used for database schema changes, while seeds refer to static data files that can be loaded into your database, neither of which contributes to the reuse of SQL transformations.