I have been toying with a new cli utility to automate some dbt development - with OpenAI, naturally, as that is how we do things nowadays. The tool parses the dbt manifest so that it can add necessary context to the tasks. It currently does the following: • Create documentation for a model • Create unit tests for a model (dbt 1.8 alpha is out!) given a description of what to test. • Generate a new model based on description and names of input models. • Alter an existing model • Explain a model • Interactively chat with a model • Fluff and rewrite a model The utility is cheesily named “dbtai”: https://github.com/radbrt/dbtai I am not able to test this with a real-world dbt project (company policies and whatnot), but if you want to give it a try I’d be happy to hear feedback. Also, someone has to have made something like this before, so feel free to post links.

Whoever tries this out, would love to know how that goes, this sounds really interesting!