‹ 首页

team-review

@nikiforovall · 收录于 昨天 · 上游提交 3 周前

Review changed code for reuse, quality, and efficiency using a team of persistent named reviewers. This skill should be used when the user says "team review", "review with team", or wants parallel code review with persistent team members for follow-up questions. Similar to /subagent-review but reviewers persist after review.

适合你,如果你想在合并代码前获得多位审查者的反馈和建议。

/ 通过 npx 安装 校验哈希
npx oh-my-skill add nikiforovall/claude-code-rules/team-review
/ 通过 bash 安装
curl -fsSL https://oh-my-skill.com/install.sh | bash -s -- nikiforovall/claude-code-rules/team-review
/ 已经装过?验证本机副本,不用重装
npx oh-my-skill verify nikiforovall/claude-code-rules/team-review
安装目标可用 --agent / --scope 或 --to 明确指定;省略时只会在唯一已存在的 agent 目录上自动选择,零命中或多命中会停止并提示。content_hash 缺失或不一致均拒装。
133GitHub stars
~939上下文体积 · 单文件
索引托管

怎么用

商店整理自技能原文 · 版本 642e21e · 表述以原文为准
它做什么

当你说出特定关键词,Claude 会创建一个三人审查团队,并行检查你的代码变更,寻找可重用性、质量和效率问题,并直接修复这些问题。审查完成后,团队成员保留,以便你后续追问。

什么时候触发

当你说“team review”、“review with team”等关键词,或者想要并行代码审查并允许后续提问时触发。

装好后可以这样说
审查后直接向特定审查员追问具体问题。
技能原文 SKILL.md作者撰写 · Apache-2.0 · 642e21e

Team Review: Code Review with Persistent Team Members

Review all changed files for reuse, quality, and efficiency using a team of named reviewers that persist for follow-up questions.

Phase 1: Identify Changes

Run git diff (or git diff HEAD if there are staged changes) to see what changed. If there are no git changes, review the most recently modified files that the user mentioned or that you edited earlier in this conversation.

Store the full diff text — you will pass it to each reviewer.

Phase 2: Create Review Team and Spawn Reviewers

Create a team (or reuse an existing one) and spawn three named team members concurrently using the Agent tool. Each member gets the full diff as context.

Use run_in_background: true for all three so they run in parallel. Give each a descriptive name parameter.

Member 1: reuse-reviewer

Prompt: You are a code reuse reviewer. Here is the diff to review:

<paste full diff>

For each change:

  1. Search for existing utilities and helpers that could replace newly written code. Look for similar patterns elsewhere in the codebase.
  2. Flag any new function that duplicates existing functionality. Suggest the existing function to use instead.
  3. Flag any inline logic that could use an existing utility — hand-rolled string manipulation, manual path handling, custom environment checks, ad-hoc type guards.

Report findings as a bulleted list. If the code is clean, say so.

Member 2: quality-reviewer

Prompt: You are a code quality reviewer. Here is the diff to review:

<paste full diff>

Review for:

  1. Redundant state: state that duplicates existing state, cached values that could be derived
  2. Parameter sprawl: adding new parameters instead of generalizing
  3. Copy-paste with slight variation: near-duplicate code blocks
  4. Leaky abstractions: exposing internal details that should be encapsulated
  5. Stringly-typed code: using raw strings where constants or enums exist
  6. Unnecessary nesting: wrapper elements that add no value
  7. Unnecessary comments: comments explaining WHAT — delete; keep only non-obvious WHY

Report findings as a bulleted list. If the code is clean, say so.

Member 3: efficiency-reviewer

Prompt: You are an efficiency reviewer. Here is the diff to review:

<paste full diff>

Review for:

  1. Unnecessary work: redundant computations, repeated file reads, duplicate API calls, N+1 patterns
  2. Missed concurrency: independent operations run sequentially when they could be parallel
  3. Hot-path bloat: blocking work added to startup or per-request hot paths
  4. Recurring no-op updates: state updates that fire unconditionally — add change-detection guards
  5. Unnecessary existence checks: pre-checking before operating (TOCTOU anti-pattern)
  6. Memory: unbounded data structures, missing cleanup, event listener leaks
  7. Overly broad operations: reading entire files when only a portion is needed

Report findings as a bulleted list. If the code is clean, say so.

Phase 3: Collect and Act on Findings

Wait for all three reviewers to complete. Aggregate their findings and fix each issue directly. If a finding is a false positive or not worth addressing, note it and move on.

When done, briefly summarize what was fixed (or confirm the code was already clean).

Important
  • Do NOT shut down the team after review — reviewers persist for follow-up questions
  • Each reviewer should use Bash(git diff), Read, Grep, Glob to explore the codebase
  • Pass the team_name parameter when spawning agents so they join the same team
  • Create tasks for each reviewer to track progress
按 Apache-2.0 许可原样转载,未经改动 · 在 GitHub 查看 →

评论

登录即可评论;带「已验证安装」的,是发布者名下有本店的安装或持有记录。