When considering the peer-review cycle there should be guidance about software citation, checking if all software that needs to be cited and is cited 'correctly' with a minimal set of important/mandatory metadata.
Code review seems like a very big step in the peer-review cycle and is not directly related to software citation.
Grant review seems like a subject on its own to determine how the peer-review cycle is established, as with code review.

Both might affect or be affected by software citation, but in my opinion, is out of scope (#80).

We currently require that a software package go through peer review before granting a DOI to a specific code release. This is a cursory review of the software artifacts accompanying a publication before it becomes citable to evaluate whether it meets some minimal standards: (a) archived in a trusted digital repository (b) can be compiled / executed (c) documented & described at an appropriate level of detail.

This depends in part on how critical the software artifacts are to the publication's findings though, we've gotten some pushback on code archival as a "who cares" type of problem.

We asked for a citable paper when accepting a code into our repository. This can be any combination of a benchmark paper, a paper on the code, or a research paper. The first indicates some validation or verification of the code has been performed and the latter that the methodology has been peer reviewed.

A lot of scholarly journals provide their reviewers with a list of questions that include things like "has the literature been adequately cited?" I don't think I've seen any in a discipline journal that ask the reviewer to consider "if software was used in this research, was it adequately cited?" Adding that one question to the review template would provide a prompt for reviewers to think about this without asking them to do a code review (which they may not feel qualified or ready to do.)

I agree with @lkellogg but don't think this is the goal of this issue - it better fits #77

the writing group thinks we need to think about citation in the context of review, but that much of review itself (such as code review) is out of scope. @danielskatz will update document

updated.