跳到主要内容
文档中心

English Documentation Chinese Cleanup Implementation Plan

For Codex: REQUIRED SUB-SKILL: Use superpowers:executing-plans to implement this plan task-by-task.

Goal: Remove all remaining Chinese text from docs/en/ while preserving technical identifiers, syntax, behavior, and consistent product terminology.

Architecture: Build an exact inventory of Chinese-containing lines, translate each passage in context using the English glossary and neighboring documentation, then repeatedly rescan until no unintended Chinese remains. Validate the final diff and documentation structure before completion.

Tech Stack: Markdown, MDX, JSON, shell validation, Docusaurus.

Task 1: Inventory residual Chinese​

Files:

  • Inspect: docs/en/**/*.md
  • Inspect: docs/en/**/*.mdx
  • Inspect: docs/en/**/*.json

Step 1: Run rg -n '[\p{Han}]' docs/en -g '*.md' -g '*.mdx' -g '*.json'.

Step 2: Group matches by file and context: prose, comments, documentation comments, strings, and example values.

Step 3: Check terminology against docs/en/glossary.md and neighboring English pages.

Task 2: Translate all human-readable Chinese​

Files:

  • Modify: Chinese-containing text files identified by Task 1.

Step 1: Translate prose, headings, labels, and descriptions.

Step 2: Translate line comments and documentation comments without changing code tokens.

Step 3: Translate human-readable strings, logs, and example values without changing APIs or syntax.

Step 4: Pause and ask the requester before changing any ambiguous behavior-sensitive string.

Task 3: Verify completeness and integrity​

Files:

  • Verify: docs/en/**

Step 1: Rerun the Chinese-character scan and require zero unexplained matches.

Step 2: Inspect git diff -- docs/en for accidental identifier, syntax, formatting, or unrelated changes.

Step 3: Check code-fence balance and parse all JSON files.

Step 4: Run the available documentation build or validation command and report unrelated pre-existing failures separately.