requesting-code-review
Dispatch a fresh-context code-reviewer subagent to catch issues before they cascade (pattern-adversarial-review, pattern-fresh-context-subagents). “Review early, review often.” The reviewer gets precisely-crafted context (a description, the requirements/plan, and the BASE→HEAD SHAs) — never the controller’s session history — so it judges the work product, not the author’s thought process → artifact-review-report.
Review is mandatory after each task in subagent-driven development, after a major feature, and before merge; optional but valuable when stuck or before refactoring. Findings are acted on by severity: fix Critical immediately, fix Important before proceeding, note Minor for later — and the skill explicitly permits pushing back with technical reasoning if the reviewer is wrong (the request side of sp-receiving-code-review). It ships the reviewer prompt template (code-reviewer.md) that sp-subagent-driven-development also uses for its final whole-branch review.
It joins the five-plus-framework code-review cluster — addy-code-review ↔ ce-code-review ↔ gstack-review ↔ bmad-code-review ↔ mp-code-review — all producing artifact-review-report; Superpowers’ distinctive angle is treating review as a dispatched fresh-context subagent gate wired into the execution loop (between every task), not a standalone command.
See Also
- sp-receiving-code-review — the reception side: how to respond to the findings this produces.
- sp-subagent-driven-development — reuses this skill’s reviewer template for the final whole-branch review.
- addy-code-review · ce-code-review · gstack-review · bmad-code-review · mp-code-review — the code-review cluster.
- artifact-review-report — the output; pattern-adversarial-review · pattern-fresh-context-subagents — the techniques.
- stage-review — the canonical stage this implements.