superpower-review

To automate code review for production readiness, dispatch a reviewer subagent using spawn_agent with specific Git SHAs. This enforces a comprehensive evaluation of code quality, architecture, and testing before merging to main.

Subagent-driven code review in LLM collaboration dispatches a dedicated reviewer subagent to evaluate changes artifact-by-artifact. It preserves your main session state while applying a structured prompt to assess production readiness.

To trigger a code review using a Git range, run spawn_agent and provide the precise task context alongside the Git SHAs. The reviewer subagent then analyzes the specified changes to generate a standardized report.

Yes, automated code review works after each task completion in subagent-driven development. It applies a defined Git range to evaluate production readiness and outputs structured recommendations before merging to main.

Yes, you can get a structured output format for code quality and architecture evaluation. The reviewer subagent enforces standardized guardrails to deliver a report containing strengths, issues, and recommendations.

A limitation of using a subagent for code review is its narrow, artifact-centric context. It focuses strictly on the code changes within the specified Git range rather than the developer's thought process.