I want a review that explains a concrete problem, because a vague second opinion only adds to my reading queue. In September 2026, I described my Claude arrangement: Sonnet implemented, Opus reviewed. I like separating those responsibilities. Adding another model to the conversation does not, by itself, establish which responsibility it received.
I am revisiting the Opus 4.7 launch in October 2026. The April announcement emphasized long programming tasks and instruction following. The practice I described months later named the Sonnet and Opus families without identifying versions. What I can bring from that experience to the announcement is an expectation for the reviewing role.
Implementation tries to produce the requested change. During review, I want its consequences examined against the expected behavior. If the second agent is merely told to continue, it may deliver more implementation. Then I have another executor and still lack the reading I wanted. The job needs a name before the model gets assigned.
In a hypothetical backend example, I want to know what happens when an input is incomplete. The reviewer should locate the relevant path and explain its conclusion with something I can check. A generic recommendation to improve robustness leaves me doing the work of discovering where a problem exists. Thanks for the essay topic; now I still need the review.
In July 2026, I considered Claude good at backend work. That confidence cannot fill in a missing requirement. If expected behavior exists only in my head, even a capable reviewer needs to infer it or ask. Making the expectation accessible is my responsibility, along with delimiting which consequences belong in this pass.
Older problems can appear during the reading. I want them identified as older so I can decide whether they belong in the current task. Comments also need a destination: accepted findings become a change or an explicit scope decision; rejected findings need an understandable reason. A loose pile of suggestions transfers arbitration to the next agent.
I will keep the roles open to different model assignments and judge reviews by the usefulness of their comments. For the next pass, I want the objective alongside the diff and findings connected to that objective. A good response should let me decide what to do next. Presenting the change again in prettier words does not justify another round.