Word counter

Skill Moosphan/word-counter-skill/skills/claude/.claude/skills/word-counter

Deterministic word-count skill for Codex and Claude Code, with consistent Chinese, English, and mixed-text counting rules.

Install

npx -y skills add Moosphan/word-counter-skill --skill word-counter

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

One thing to look at

1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Deterministic counting for Chinese, English, and mixed-language text such as articles, essays, novels, chapters, outlines, transcripts, and pasted paragraphs. Use when the user asks for word count, character count, manuscript length, bilingual text count, chapter length, or wants a stable count instead of a model estimate.

SKILL.md

2.1 KB, as published. Nobody here has run it

Word Counter

Use the bundled script to produce a reproducible count instead of estimating from the model.

If the user asks in Chinese, run the script with --locale zh so the report is returned in Chinese. If the user asks in English, use --locale en. Only use --details when the user explicitly asks for detailed statistics.

Pick A Profile

Use zh for Chinese-style content counting.
Use en for English-style word counting.
Use mixed for one total across Chinese and English text.

If the user is ambiguous, run mixed first and mention the zh and en alternatives.

Run The Script

For pasted content:

python3 "$HOME/.claude/skills/word-counter/scripts/word_counter.py" --profile mixed --locale en --format markdown --text "Hello world from OpenAI"

For a file:

python3 "$HOME/.claude/skills/word-counter/scripts/word_counter.py" --profile zh --locale zh --format markdown ./chapter-01.txt

Return A Clear Summary

Return this compact Markdown structure. Use the English example as the default format reference:

# Word Count Result

- Selected profile: `mixed` (Mixed-language count)
- Selected total: `1234`
- Applied formula: `mixed_count = cjk_chars + english_words + number_tokens + other_words`

| Metric | Value |
| --- | ---: |
| Chinese count | 1300 |
| English words | 45 |
| Mixed total | 1234 |
| Line count | 12 |
| Paragraph count | 4 |

For Chinese replies, use the same compact structure with Chinese labels. Do not show CJK characters, English words, or Number tokens unless the user explicitly asks for detailed statistics, in which case run with --details.

For rules and examples, read references/counting-rules.md.