requesting-code-review

You can automate a pre-merge code review by dispatching a code-reviewer subagent with your implementation summary, plan reference, and precise BASE_SHA..HEAD_SHA git range values to receive categorized and actionable feedback.

A structured code review process uses a checklist-driven template to evaluate code changes and categorize identified defects by severity—Critical, Important, or Minor—while requiring specific file:line references and remediation guidance for each finding.

You should request a code review after each completed task in subagent-driven development, after major feature completion, before merging branches, when feeling stuck on an implementation, or prior to starting large refactoring efforts.

To run a code review, you must provide a concise summary of what was implemented, a plan or requirements reference, and the precise BASE_SHA and HEAD_SHA git commit values defining the scope of the changes.

No, the code review subagent relies on a checklist-driven review template to ensure a repeatable process that systematically prevents regressions and misaligned implementations before they compound or reach production.

A structured code review prevents regressions and misaligned implementations by providing a repeatable, checklist-driven process that identifies critical, important, and minor issues before they compound or reach production environments.