跳到技能内容
词映 AI导航
发现工具选型指南Skills 技能AI 快讯AI图片AI视频AI文档AI办公AI助手AI搜索AI编程AI音频AI设计AI智能体模型与开发AI学习模型评测AI提示词内容检测AI金融AI 3D扫描与识别其他工具跨境电商 AI
词映 AI导航
工具库Skills 技能
我的收藏
词映 AI导航
发现工具选型指南Skills 技能AI 快讯AI图片AI视频AI文档AI办公AI助手AI搜索AI编程AI音频AI设计AI智能体模型与开发AI学习模型评测AI提示词内容检测AI金融AI 3D扫描与识别其他工具跨境电商 AI
Skills directory/stumman/hermes-harness-skills/critical-review

Codex critical-review skill setup and severity-ranked code review

Review a change against its acceptance criteria and report evidence-based findings across correctness, design, readability, and tests. Group findings by severity with a location, impact, and suggested fix.

中文

Install in Codex

A Codex project workspace. The sample includes the changed expression, acceptance criterion, and observation for a severity-ranked code review.

  1. Review `.agents/skills/critical-review/SKILL.md` and its requirements at the pinned GitHub commit.
  2. Create `.agents/skills/critical-review/` in your project and save the reviewed file as `.agents/skills/critical-review/SKILL.md`.
  3. Start Codex from that project's root, then type `$critical-review` and select the matching skill.

Example task

Invoke critical-review to review this fictional order totals change.
case data `order_totals_test` | acceptance=sum all items
file `order_totals.py` | changed expression `sum(items[1:])`
input `items=[3]` | observed `total=0` | expected `total=3`

What to check

Report the skipped first item in `order_totals.py` as a `Blocker`. For `items=[3]`, the observation is `total=0` and acceptance requires `total=3`. Suggest fixing the sum range and keeping this boundary case.

Original SKILL.md on GitHub · Updated 2026-11-14

Frequently asked question

How do I make a review finding actionable?

Group findings by severity and give the changed location, the impact, and a verifiable fix. The sample affects the correctness of the order total.

Browse Codex skills in the Chinese directory
词映 AI 导航Skills 技能
工具分类全部工具已本地化技能选型指南AI 快讯