agentsclimarketplace

Requesting code review

Skill jnMetaCode/superpowers-zh/skills/requesting-code-review

完成任务、实现重要功能或合并前使用,用于验证工作成果是否符合要求From its SKILL.md

Install
npx -y skills add jnMetaCode/superpowers-zh --skill requesting-code-review

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

What its file declares

Copied from the file, not written here

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

2.9 KB, as published. Nobody here has run it

请求代码审查

派遣代码审查子代理,在问题扩散之前发现它们。审查者获得的是精心组织的评估上下文——绝不是你的会话历史。

核心原则: 早审查,勤审查。

何时请求审查

必须审查:

  • 子代理驱动开发中每个任务完成后
  • 完成重要功能后
  • 合并到 main 之前

可选但有价值:

  • 卡住时(换个视角)
  • 重构之前(建立基线)
  • 修复复杂 bug 之后

如何请求

1. 获取 git SHA:

BASE_SHA=$(git rev-parse HEAD~1)  # 或 origin/main
HEAD_SHA=$(git rev-parse HEAD)

2. 派遣代码审查子代理:

使用 Task 工具,指定 general-purpose 类型,填写 code-reviewer.md 中的模板

占位符说明:

  • {DESCRIPTION} - 你刚完成的内容简要说明
  • {PLAN_OR_REQUIREMENTS} - 预期功能
  • {BASE_SHA} - 起始提交
  • {HEAD_SHA} - 结束提交

3. 处理反馈:

  • Critical 问题立即修复
  • Important 问题在继续之前修复
  • Minor 问题记录下来稍后处理
  • 如果审查者有误,用技术理由反驳

示例

[刚完成任务 2:添加验证功能]

你:让我在继续之前请求代码审查。

BASE_SHA=$(git log --oneline | grep "Task 1" | head -1 | awk '{print $1}')
HEAD_SHA=$(git rev-parse HEAD)

[派遣代码审查子代理]
  DESCRIPTION: 添加了 verifyIndex() 和 repairIndex(),支持 4 种问题类型
  PLAN_OR_REQUIREMENTS: docs/superpowers/plans/deployment-plan.md 中的任务 2
  BASE_SHA: a7981ec
  HEAD_SHA: 3df7661

[子代理返回]:
  优点:架构清晰,测试真实
  问题:
    Important:缺少进度指示器
    Minor:报告间隔使用了魔法数字 (100)
  评估:可以继续

你:[修复进度指示器]
[继续任务 3]

常见的合理化借口

借口现实
"我自己看一下 diff 就行了,不用专门派审查者"你是协调者——在自己的会话里读 diff 会烧掉你继续推进工作所需的上下文窗口。派一个审查子智能体:diff 和评估过程都待在它的上下文里,只有结论回到你这里。
"审查者需要我的全部会话历史才能理解这次改动"给它精心组织的上下文,绝不给会话历史。这样审查者才会盯着工作成果,而不是你的思考过程。

红线

绝不要:

  • 因为"很简单"就跳过审查
  • 忽略 Critical 问题
  • 带着未修复的 Important 问题继续推进
  • 对合理的技术反馈进行争辩

如果审查者有误:

  • 用技术理由反驳
  • 展示证明其可行的代码/测试
  • 要求澄清

参见模板:requesting-code-review/code-reviewer.md

What ships with it: 1 file

4.5 KB alongside SKILL.md

Gives 4 of the 12 instructions most code review skills give

Counted across 668 of the 814 authors here whose files we hold, read 2026-09-06

  • Provide technical reasoning when pushing backin 84 of 668, across 70 files
  • Fix critical issues immediatelyhere, and in 77 of 668, across 60 files
  • Dispatch a code reviewer subagenthere, and in 76 of 668, across 59 files
  • Fix important issues before proceedinghere, and in 73 of 668, across 56 files
  • Ask for clarification on unclear itemsin 68 of 668, across 56 files
  • Verify feedback against codebase before implementationin 66 of 668, across 55 files
  • Implement fixes one at a timein 64 of 668, across 53 files
  • Test each fix individuallyin 62 of 668, across 51 files
  • Restate technical requirements in own wordsin 57 of 668, across 46 files
  • Reply to inline comments in the specific threadin 51 of 668, across 40 files
  • Note minor issues for laterhere, and in 49 of 668, across 34 files
  • Group findings by severityin 48 of 668, across 47 files

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once.

Keep looking

Skills are one crate of 626,194. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.