Finance wants peer review to catch both notebook logic and job orchestration changes before merge. What should reviewers examine together?

Finance peer review before merge must catch both notebook transformation logic and job orchestration changes, so reviewers should examine pull-request diffs that include job YAML and notebook changes as one review scope. Civic pipelines fail in production when notebooks are correct but schedules, tasks, or parameters are wrong—or the reverse—so treating orchestration and code as a single review unit is the sound practice. Pull-request diffs that include both job YAML and notebook outputs as one review scope fails because notebook outputs show prior results, not the notebook transformation code, so reviewers cannot catch logic changes that the orchestration depends on. Pull-request diffs that include both job run logs and notebook changes as one review scope fails because job run logs show past execution history, not job orchestration like schedules, tasks, or parameters that must be seen with the notebook logic. Pull-request diffs that include both job YAML and notebook permissions as one review scope fails because notebook permissions address access control, not the notebook transformation logic that must be reviewed together with the job YAML. Exam caveat: DP-750-style governance expects reviewers to see orchestration and transform assets together in Git-based PRs. Operational check: configure PR templates or CODEOWNERS so job YAML and notebooks appear together, and reject merges where reviewers signed off without opening both classes of diffs.