code-review

Dispatch the code-reviewer subagent with SHA ranges and run verification gates.

3|Updated Dec 28, 2025
One-click install
npx skills add https://github.com/hongbietcode/synthetic-claude --skill code-review-hongbietcode
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-review
Source: https://github.com/hongbietcode/synthetic-claude/tree/main/plugins/basic-workflow/skills/code-review
Command: npx skills add https://github.com/hongbietcode/synthetic-claude --skill code-review-hongbietcode

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Code review is essential to ensure technical rigor and prevent premature completion. This Skill codifies best practices for handling feedback, requesting formal reviews via a code-reviewer subagent, and applying verification gates before any status claims.

Core Features & Use Cases

  • Receiving Feedback Protocol: READ → UNDERSTAND → VERIFY → EVALUATE → RESPOND → IMPLEMENT to evaluate and act on feedback without performative agreement.
  • Requesting Review Protocol: After each task or major feature, dispatch the code-reviewer subagent with the required SHAs and summary to obtain expert input before proceeding.
  • Verification Gates: Run verification commands (tests/build/status) and claim success only with fresh evidence.

Quick Start

Use the code-review skill after completing a task: dispatch the code-reviewer subagent with BASE_SHA and HEAD_SHA, then run verification gates before claiming completion.

Frequently Asked Questions about code-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate rigorous code reviews before merging a pull request?

Automated code reviews use a code-reviewer subagent and verification gates to prevent premature completion. You dispatch the subagent with BASE_SHA and HEAD_SHA to enforce evidence-based technical rigor before merging.

What is the best way to handle code review feedback without performative agreement?

Handling code review feedback involves a READ → UNDERSTAND → VERIFY → EVALUATE → RESPOND → IMPLEMENT protocol. This ensures you evaluate feedback technically and verify claims before acting, preventing performative agreement.

When do I need to run verification gates during software development?

Run verification gates after completing each task, finishing major features, and before merging. You must execute tests, builds, or status commands to claim success only with fresh evidence.

Can I use a subagent to prevent premature completion of coding tasks?

Yes, dispatching a code-reviewer subagent enforces technical rigor and prevents premature completion. It requires technical justification and clarifying questions before allowing any status claim.

Does the code review process work without dependencies?

Yes, the code review process works without external dependencies. It uses an internal code-reviewer subagent and verification gates to guide evidence-based reviews natively.

Why does claiming task success require fresh evidence?

Claiming success requires fresh evidence to enforce technical rigor and prevent premature completion. Verification gates ensure success claims are backed by recently executed tests, builds, or status checks.