Back to Details

code-review

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Code Review

代码评审

Dispatch specialized review agents matched to the situation. Choose a mode or let context guide you.
根据场景调配专业的评审Agent。可自行选择模式，也可根据上下文自动适配。

Review Modes

评审模式

Quick Review

快速评审

When: Ad-hoc checks, quick sanity check, low-noise feedback Dispatch: quick-reviewer agent (confidence scoring, reports only >= 80) Output: High-confidence issues grouped by severity (Critical 90-100, Important 80-89)
适用场景： 临时检查、快速合理性验证、低干扰反馈 调配Agent： quick-reviewer agent（置信度评分，仅报告评分≥80的问题） 输出结果： 按严重程度分组的高置信度问题（严重：90-100，重要：80-89）

Deep Review

深度评审

When: After completing a feature, before merge, in subagent-driven-development loop Dispatch: deep-reviewer agent (full checklist with template placeholders) Placeholders: {WHAT_WAS_IMPLEMENTED}, {PLAN_OR_REQUIREMENTS}, {BASE_SHA}, {HEAD_SHA}, {DESCRIPTION} Output: Strengths, Issues (Critical/Important/Minor), Recommendations, Merge verdict
适用场景： 功能开发完成后、合并前、子Agent驱动的开发循环中 调配Agent： deep-reviewer agent（带模板占位符的完整检查清单） 占位符： {WHAT_WAS_IMPLEMENTED}, {PLAN_OR_REQUIREMENTS}, {BASE_SHA}, {HEAD_SHA}, {DESCRIPTION} 输出结果： 优势分析、问题（严重/重要/次要）、改进建议、合并判定结果

Specialist Review

专项评审

When: Targeted analysis of a specific quality dimension Dispatch: One or more specialist agents:
AgentFocus
comment-analyzerComment accuracy, documentation quality, comment rot
test-analyzerTest coverage quality, behavioral coverage gaps, test resilience
silent-failure-hunterError handling, silent failures, catch block specificity
type-design-analyzerType invariants, encapsulation, type usefulness
适用场景： 针对特定质量维度的定向分析 调配Agent： 一个或多个专项Agent：
Agent关注重点
comment-analyzer注释准确性、文档质量、注释过时
test-analyzer测试覆盖质量、行为覆盖缺口、测试韧性
silent-failure-hunter错误处理、静默故障、捕获块针对性
type-design-analyzer类型不变性、封装性、类型实用性

Comprehensive Review

全面评审

When: Before merging to main, finishing a development branch Dispatch: 4-pass structure (TDD compliance, diff-only quality, contextual quality, browser audit). See references/comprehensive-review-procedure.md for pass details and ordering.
适用场景： 合并至主分支前、开发分支完成时 调配结构： 4轮检查（TDD合规性、仅差异质量检查、上下文质量检查、浏览器审计）。有关各轮检查的细节和顺序，请查看references/comprehensive-review-procedure.md。

How to Request

如何发起评审

  1. Gather git context using 
    scripts/review-prep
    (auto-detects base branch, outputs BASE_SHA, HEAD_SHA, diff stats, changed files, commit log). Use 
    --json
    for structured output.
  2. Read the reference file for your chosen agent(s)
  3. Spawn 
    general-purpose
    Task agent with the prompt + review-prep output
  4. Act on feedback: fix Critical immediately, fix Important before proceeding
  1. 使用
    scripts/review-prep
    收集git上下文信息（自动检测基础分支，输出BASE_SHA、HEAD_SHA、差异统计、变更文件、提交日志）。使用
    --json
    参数可获取结构化输出。
  2. 阅读所选Agent的参考文件
  3. 结合提示词和review-prep输出，启动
    general-purpose
    任务Agent
  4. 根据反馈采取行动：立即修复严重问题，在继续开发前修复重要问题

Workflow Integration

工作流集成

WorkflowReview Mode
subagent-driven-development (per-task)Deep Review
executing-plans (batch)Deep Review
Before merge to mainComprehensive Review
Ad-hoc / when stuckQuick Review
Specific concernSpecialist Review
GitHub PR review + postinggithub-pr skill (automated PR review orchestrator)
For code simplification, see the code-simplify skill.
工作流评审模式
subagent-driven-development（按任务）深度评审
executing-plans（批量）深度评审
合并至主分支前全面评审
临时检查/遇到瓶颈时快速评审
特定关注点专项评审
GitHub PR评审+发布github-pr skill（自动化PR评审编排器）
如需代码简化，请查看code-simplify skill。

Red Flags

警示信号

  • Skipping review because "it's simple"
  • Ignoring Critical issues
  • Proceeding with unfixed Important issues
  • Running Comprehensive passes all sequentially (1+2 are independent, parallelize)
  • Not deduplicating between Pass 2 and Pass 3 in Comprehensive mode
  • 以“代码很简单”为由跳过评审
  • 忽略严重问题
  • 未修复重要问题就继续开发
  • 按顺序执行全面评审的所有轮次（第1、2轮是独立的，可并行执行）
  • 在全面评审模式下未对第2轮和第3轮的结果进行去重

Reference Files

参考文件

AgentReference
deep-reviewerreferences/deep-reviewer.md
quick-reviewerreferences/quick-reviewer.md
comment-analyzerreferences/comment-analyzer.md
test-analyzerreferences/test-analyzer.md
silent-failure-hunterreferences/silent-failure-hunter.md
type-design-analyzerreferences/type-design-analyzer.md
Agent参考文档
deep-reviewerreferences/deep-reviewer.md
quick-reviewerreferences/quick-reviewer.md
comment-analyzerreferences/comment-analyzer.md
test-analyzerreferences/test-analyzer.md
silent-failure-hunterreferences/silent-failure-hunter.md
type-design-analyzerreferences/type-design-analyzer.md