Menu

#526 by GitHub stars

coji/natural-japanese

仕事の日本語文書を、自然で読みやすく、わかりやすい日本語で書く・直すためのスキル。議事録、調査レポート、顧客向けのレポート、社内ガイド、技術記事、PR の説明文、リサーチメモ、企画書、メール、スライド構成などの作成や推敲と、note やブログ、エッセイの執筆に使う。文字起こしから議事録を作るときにも使う。自然な日本語にしたい、読みやすくしたい、簡潔にしたい、結論から書きたいといった依頼のほか、読みづらい、わかりにくい、何を言っているかわからない、AI っぽい、AI 臭い、英語の直訳っぽい、日本語として不自然、違和感があるといった指摘にも使う。書き換えずに AI らしさを判定・採点してほしいとい

Claude CodeCodexWriting

History

Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.

View weekly changes

skills/natural-japanese/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Local scripts → After Local scripts

25a8638b7800 → 25a8638b7800

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Local scripts

First observation → 25a8638b7800

No earlier version was observed for comparison.

Compare on GitHub

View on GitHubAdd to my skills

Verified commit 25a8638b780037a2012e2bb3f624a0b1710666fb

git clone --filter=blob:none --no-checkout https://github.com/coji/natural-japanese.git .hotskills/coji-natural-japanese && git -C .hotskills/coji-natural-japanese checkout --detach 25a8638b780037a2012e2bb3f624a0b1710666fb # Copy each selected skill directory from .hotskills/coji-natural-japanese into .claude/skills/<skill-name>/
git clone --filter=blob:none --no-checkout https://github.com/coji/natural-japanese.git .hotskills/coji-natural-japanese && git -C .hotskills/coji-natural-japanese checkout --detach 25a8638b780037a2012e2bb3f624a0b1710666fb # Copy each selected skill directory from .hotskills/coji-natural-japanese into .agents/skills/<skill-name>/
2kGitHub stars
24stars in 7 days
1skill in this repository
183commits in 30 days

Heat

7.4
  • Star growth, last 7 daysAhead of 87% of tracked repositories
  • Star growth, last 24 hoursAhead of 93% of tracked repositories
  • Rank movement, last 7 daysAhead of 64% of tracked repositories
  • Commits, last 30 daysAhead of 96% of tracked repositories
  • Accounts asking about it through MCP, last 7 daysNot available
  • Hot votes this weekAhead of 0% of tracked repositories

Heat never changes the order by stars, and sponsorship never touches it. How heat works

Stars over time

Aug 2, 2026Nov 6, 20262,023

Individual skills in this repository

How it works

natural-japanese

When your agent uses it

仕事の日本語文書を、自然で読みやすく、わかりやすい日本語で書く・直すためのスキル。議事録、調査レポート、顧客向けのレポート、社内ガイド、技術記事、PR の説明文、リサーチメモ、企画書、メール、スライド構成などの作成や推敲と、note やブログ、エッセイの執筆に使う。文字起こしから議事録を作るときにも使う。自然な日本語にしたい、読みやすくしたい、簡潔にしたい、結論から書きたいといった依頼のほか、読みづらい、わかりにくい、何を言っているかわからない、AI っぽい、AI 臭い、英語の直訳っぽい、日本語として不自然、違和感があるといった指摘にも使う。書き換えずに AI らしさを判定・採点してほしいという依頼や、自分の文体を学ばせたいという依頼にも対応する。

What it loads into context

Frontmatter

308 tokens

Always available to the agent

SKILL.md

14,716 tokens

When the skill is activated

Reference documents

22,182 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

  1. いちばん大事な規則:元にない内容を足さない
  2. 実行モード
  3. 呼び出し方
  4. 1. 書く前に決める
  5. 2. 書くときの原則
  6. 3. 既存の文書を直すとき
  7. 4. lint で機械的に検出する
  8. 5. 構造と読みやすさを点検する
  9. 6. 収束と仕上げ
  10. 7. 後片付け

Scripts executed separately

  • scripts/calibrate.py
  • scripts/calques.py
  • scripts/lint.py
  • scripts/newfacts.py
  • scripts/outline.py
  • scripts/paragraphs.py
  • scripts/semantic.py
  • scripts/terms.py
  • scripts/textcore.py

Files

  • ../../.claude-plugin/marketplace.jsonasset616 bytes
  • ../../.claude-plugin/plugin.jsonasset610 bytes
  • SKILL.mdcore51,886 bytes
  • SKILL.mdcore51,886 bytes
  • assets/style-profile-template.mddoc2,318 bytes
  • references/diagnose.mddoc5,032 bytes
  • references/doctypes/guide.mddoc14,289 bytes
  • references/doctypes/memo.mddoc10,995 bytes
  • references/doctypes/minutes.mddoc8,902 bytes
  • references/doctypes/report.mddoc11,618 bytes
  • references/doctypes/slide.mddoc9,221 bytes
  • references/manual-checklist.mddoc7,928 bytes
  • scripts/calibrate.pyruns33,586 bytes
  • scripts/calques.pyruns5,520 bytes
  • scripts/fixtures/ai-smelly.mddoc3,978 bytes
  • scripts/fixtures/natural.mddoc2,827 bytes
  • scripts/lint.pyruns134,524 bytes
  • scripts/newfacts.pyruns6,349 bytes
  • scripts/outline.pyruns18,817 bytes
  • scripts/paragraphs.pyruns5,112 bytes
  • scripts/semantic.pyruns17,511 bytes
  • scripts/terms.pyruns12,466 bytes
  • scripts/textcore.pyruns23,524 bytes
Read SKILL.md

# natural-japanese

仕事の日本語文書を、読みやすくわかりやすく書く・直すためのスキル。議事録・調査レポート・社内ガイド・リサーチメモ・スライドといった仕事の文書から、技術記事・PR の説明文・note・ブログ・エッセイまでを扱う。AI らしい書き癖を取り除くことも含む。

疑わしい箇所の検出は `scripts/lint.py` が機械的に行い、直すかどうかはあなたが文脈を見て決める。AI は自分の書き癖に気づきにくいので、検出を機械に任せている。

## いちばん大事な規則:元にない内容を足さない

既存の文書を直すときは、元の文書に書かれていない事実・数値・人名・日付・条件・評価・意見・方針を足さない。直した文書を読んだ人は、そこに書かれていることを書き手が言ったこととして受け取る。足した内容は、たとえ正しくても、書き手が言っていないことを書き手の言葉として広めてしまう。

次の変更も「足す」に当たる。

- 確信の度合いを変える。「次回また相談する」を「合意した」にする、「かもしれない」を「決めた」にする、逆に断定を推定に弱める
- 書き手の経験や見解を書き足す。「これには何度も助けられた」「筆者としては〜をすすめる」など
View the complete source on GitHub

Skills included

  • natural-japanese

Repository facts

License
MIT
Created
Jul 12, 2026
Last pushed
Nov 5, 2026
Forks
58
Open issues
0
Project activity
Active

Rank over the last 12 weeks

Now #526 · best #443 · down 61 in 30 days

  1. #524 from Oct 30 to Oct 31
  2. #526 on Nov 1
  3. #529 on Nov 2
  4. #528 on Nov 3
  5. #529 on Nov 4
  6. #526 since Nov 5
78 earlier periods
  1. #870 on Aug 2
  2. #863 on Aug 7
  3. #872 on Aug 8
  4. #874 on Aug 9
  5. #909 from Aug 11 to Aug 12
  6. #890 on Aug 14
  7. #858 on Aug 15
  8. #876 on Aug 16
  9. #860 on Aug 17
  10. #872 on Aug 18
  11. #857 on Aug 19
  12. #845 on Aug 20
  13. #853 on Aug 21
  14. #799 on Aug 22
  15. #855 on Aug 23
  16. #865 on Aug 24
  17. #875 on Aug 25
  18. #860 on Aug 26
  19. #867 on Aug 27
  20. #903 on Aug 28
  21. #818 on Aug 29
  22. #824 on Aug 30
  23. #867 on Aug 31
  24. #898 on Sep 1
  25. #712 on Sep 2
  26. #650 on Sep 3
  27. #616 on Sep 4
  28. #609 on Sep 5
  29. #616 on Sep 6
  30. #627 on Sep 7
  31. #640 on Sep 8
  32. #626 on Sep 9
  33. #630 on Sep 10
  34. #609 on Sep 11
  35. #603 on Sep 12
  36. #594 on Sep 13
  37. #611 on Sep 14
  38. #603 on Sep 15
  39. #613 on Sep 16
  40. #611 from Sep 17 to Sep 18
  41. #588 on Sep 19
  42. #599 on Sep 20
  43. #606 on Sep 21
  44. #552 on Sep 22
  45. #499 on Sep 23
  46. #471 on Sep 24
  47. #470 on Sep 25
  48. #462 on Sep 26
  49. #475 on Sep 27
  50. #478 on Sep 28
  51. #482 on Sep 29
  52. #478 on Sep 30
  53. #471 on Oct 1
  54. #463 on Oct 2
  55. #448 on Oct 3
  56. #443 on Oct 4
  57. #452 on Oct 5
  58. #459 on Oct 6
  59. #465 on Oct 7
  60. #476 on Oct 8
  61. #478 on Oct 9
  62. #472 on Oct 10
  63. #475 on Oct 11
  64. #485 on Oct 12
  65. #482 on Oct 13
  66. #485 on Oct 14
  67. #489 on Oct 15
  68. #491 on Oct 16
  69. #480 on Oct 17
  70. #496 on Oct 19
  71. #500 on Oct 20
  72. #504 on Oct 21
  73. #498 on Oct 22
  74. #491 on Oct 23
  75. #480 on Oct 24
  76. #522 on Oct 25
  77. #520 on Oct 28
  78. #523 on Oct 29

Before you install

Local scripts

Automated signals, not a guarantee

Checked Nov 3, 2026

Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.

Show 1 note
  • skills/natural-japanese/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
Show all 13 signals by file

.claude-plugin/plugin.json

  • Network accessline 6
    "name": "coji"   },   "homepage": "https://github.com/coji/natural-japanese",
  • Network accessline 7
    },   "homepage": "https://github.com/coji/natural-japanese",   "repository": "https://github.com/coji/natural-japanese",
  • Network accessline 8
    "homepage": "https://github.com/coji/natural-japanese",
  • Network accessline 9
    "repository": "https://github.com/coji/natural-japanese",

skills/natural-japanese/scripts/calibrate.py

  • Executable scriptline 1
    Executable file

skills/natural-japanese/scripts/calques.py

  • Executable scriptline 1
    Executable file

skills/natural-japanese/scripts/lint.py

  • Executable scriptline 1
    Executable file

skills/natural-japanese/scripts/newfacts.py

  • Executable scriptline 1
    Executable file

skills/natural-japanese/scripts/outline.py

  • Executable scriptline 1
    Executable file

skills/natural-japanese/scripts/paragraphs.py

  • Executable scriptline 1
    Executable file

skills/natural-japanese/scripts/semantic.py

  • Executable scriptline 1
    Executable file

skills/natural-japanese/scripts/terms.py

  • Executable scriptline 1
    Executable file

skills/natural-japanese/scripts/textcore.py

  • Executable scriptline 1
    Executable file

How the signals fit together

Facts from a static read of the files at the pinned commit. They describe what the code does, not why.

Some files could not be read; the facts below still hold.

Signal confidence: 9 facts

Works with

Agent compatibility, from the files in this repository
AgentStatusEvidenceInstall to
Claude CodeNativeManifest .claude-plugin/marketplace.json.claude/skills/<skill-name>/
CodexNativeManifest AGENTS.md.agents/skills/<skill-name>/
CursorCompatibleStandard SKILL.md skills/natural-japanese/SKILL.md.cursor/skills/<skill-name>/
Gemini CLICompatibleStandard SKILL.md skills/natural-japanese/SKILL.md.gemini/skills/<skill-name>/
GitHub CopilotCompatibleStandard SKILL.md skills/natural-japanese/SKILL.md.github/skills/<skill-name>/
OpenCodeCompatibleStandard SKILL.md skills/natural-japanese/SKILL.md.opencode/skills/<skill-name>/
  • Native The repository ships a manifest, folder or frontmatter field for this agent.
  • Compatible A standard SKILL.md that this agent can read.
  • Unknown No evidence either way.

Derived automatically from paths, manifests and frontmatter. Not tested on each agent.

You might also use

nanaism/yomiyasu

AIが生成した日本語を推敲するスキル。利用者が yomiyasu の名前を出して頼んだとき（「/yomiyasu」、「$yomiyasu:yomiyasu」、「yomiyasu で推敲して」など）と、スキルを使った日本語の推敲…

2.9kGitHub stars

epoko77-ai/im-not-ai

AI가 쓴 한글을 사람 글처럼 윤문하는 Claude 스킬 — Korean AI-text humanizer: detects and rewrites translationese, mechanical parallelism, and 71 other AI tells

6.4kGitHub stars

skillmds/skillmd

Lint, search, and install Agent Skills (SKILL.md) — CLI + MCP server + lint engine

189GitHub stars

arslanrakh/write-human

Offline linter that marks LLM-sounding prose. Its statistical model is trained on 2.8B tokens of human text and 11 LLM families, with 0.96 mean AUROC on 15 held-out sets and no mo…

13GitHub stars

AshwinSathian/humanize-writing-skill

Claude Code skill that makes Claude's prose read as written by a person. Rules on claims, sentence shape, and endings from 2026 research, tested blind on Haiku, Sonnet, and Opus. …

2GitHub stars

conorbronsdon/avoid-ai-writing

Skill that audits and rewrites content to remove AI writing patterns. Use it with your favorite agents including Claude Code, OpenClaw, Codex, and Hermes.

5.5kGitHub stars

Explore alternatives and comparisons

Add this badge to your README

It updates every night and links to this page.

coji/natural-japanese: #526 on hotskills, 2k stars

[![coji/natural-japanese on hotskills](https://hotskills.dev/badge/coji-natural-japanese.svg)](https://hotskills.dev/skill/coji-natural-japanese/)