Learning AI-Native Software Engineering

Chapter 6. The SDD Workflow

A Note for Early Release Readers

With Early Release ebooks, you get books in their earliest form—the author’s raw and unedited content as they write—so you can take advantage of these technologies long before the official release of these titles.

This will be the 7th chapter of the final book. Please note that the GitHub repo will be made active later on.

If you’d like to be actively involved in reviewing and commenting on this draft, please reach out to the editor at [e-mail].

[[Link to Come]] made the case for why writing specifications matters. This chapter is about what happens next: the full workflow that takes a specification from intent to production-ready code, then keeps that code aligned with the spec over time. Along the way, it also covers what makes a spec agent-friendly in practice, so that the workflow has good inputs to work with at every phase. It’s the operational heart of SDD.

The SDD workflow isn’t a new process layer bolted on top of how you already work. It’s a set of named phases with durable artifacts, human checkpoints at the transitions between phases, and clear rules for when to continue and stop. Once you’ve run the loop a few times, the structure becomes lightweight. The overhead is low; the discipline is what makes the difference.

This chapter also addresses one of the most common questions engineers raise when they start exploring SDD: “This sounds great for greenfield projects, but what about the codebase I’m ...