requesting-code-review

You can automate code review by providing a BASE_SHA and HEAD_SHA range to scope the exact changes. The reviewer subagent then evaluates only those commits, ensuring focused feedback on the specific code modifications before they progress.

Pre-merge checks are best handled by dispatching a focused code reviewer subagent with task-specific context. This validates feature completion and identifies Critical, Important, or Minor issues before the code integrates into the main branch.

Code quality assurance in subagent-driven workflows is handled by dispatching a dedicated reviewer subagent at development checkpoints. It uses tightly crafted context and SHA-based scoping to provide prioritized, actionable remediation feedback.

No, you do not need the full chat history. The code review process uses tightly crafted, task-specific context along with BASE_SHA and HEAD_SHA inputs to evaluate changes, preventing defects and requirement mismatches without full conversation logs.

You should request a code review after completing a subagent-driven development task or finishing a major feature. Applying these pre-merge validation checkpoints catches integration risks early and reduces late-stage rework.

Code review feedback is prioritized into Critical, Important, and Minor categories. This prioritization-driven remediation guides you to fix the most severe defects and requirement mismatches first before continuing with your work.