Requesting Code Review

What problem does it solve?

Waiting for human code reviews can slow down development, and manual reviews might miss subtle issues. This skill automates the process of requesting a code review from a dedicated AI subagent, providing immediate, structured feedback against a plan or requirements, ensuring quality and accelerating iteration.

Core Features & Use Cases

• Automated Subagent Dispatch: Dispatches a specialized code-reviewer subagent with the necessary context (code changes, plan, SHAs).
• Structured Review Template: Uses a detailed template to guide the subagent's review across code quality, architecture, testing, and requirements.
• Early Issue Detection: Integrates review checkpoints into workflows like subagent-driven development to catch issues before they compound.
• Use Case: After completing a task in your development plan, you'll use this skill to trigger an AI code review. The subagent will analyze your changes against the plan and provide actionable feedback, allowing you to fix issues immediately.

Quick Start

Get Git SHAs for the changes

BASE_SHA=$(git rev-parse HEAD~1) HEAD_SHA=$(git rev-parse HEAD)

Dispatch code-reviewer subagent (example for Claude Code / AI environment)

Task("Code Review", { WHAT_WAS_IMPLEMENTED: "Verification and repair functions for conversation index", PLAN_OR_REQUIREMENTS: "Task 2 from docs/plans/deployment-plan.md", BASE_SHA: BASE_SHA, HEAD_SHA: HEAD_SHA, DESCRIPTION: "Added verifyIndex() and repairIndex() with 4 issue types" })