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/ToaruPen/coq-japanese_stable/qudjp-function-word-residue-audit

Codex QudJP Audit for Stray English in Japanese Text

Use runtime evidence and static candidates to distinguish stray English in localized output from stale test expectations and deliberate English.

中文

Install in Codex

Codex, the pinned QudJP checkout, Python, `just`, and an accessible local test log.

  1. Copy the pinned `.codex/skills/qudjp-function-word-residue-audit/` directory into `.agents/skills/` in the project.
  2. Select the skill and point it at a local `Player.log` or a defined static audit scope.

Example task

In a practice log, leave `to the east` in a Japanese location message. Ask which route generated the phrase.

What to check

Expect reproduction evidence, the owning layer, and the test to update. Deliberate hotkeys and names are classified separately.

Original SKILL.md on GitHub · Updated 2026-11-13

Frequently asked question

Should every English fragment be translated?

Distinguish accidental grammar fragments from hotkeys, names, and intentionally retained text.

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