This document provides guidelines for authors and reviewers of YANG module documents. It obsoletes RFC 8407 and also updates RFC 8126.
I think that it is clear and well-written.

However, I have few suggestions:

- I would include in section 1 more information about the motivations behind the changes proposed in the document. Some are due to errors, others are consequential to the YANG implementation experience, and so on. Maybe the long list of section 1.1 can be split into categories. It is just to provide additional context for readers.

- In section 2.4, the meaning of the uppercase usage of the key words could be further explained. Since this document provides guidelines for YANG Data Models, I think that a sentence to clarify the implications of the normative terminology would help in this case. For example, if the normative terminology is needed to establish the level of compliance of every IETF YANG Data Models with these guidelines, it is good to highlight this point in section 2.

- I would point out in section 3.5.1 that, in addition to service, network and device models, other types of YANG modules are possible and have been defined covering layering relationships, e.g. between underlay networks and overlay services.

- I'm wondering whether it can be useful in section 4 to provide some recommendations about the typical structure and ordering while writing a YANG data model.