Agent skill

Humanizer Zh

by ai-zixun in ai-zixun/humanizer-zh

Remove signs of AI-generated, translated, or overly mechanical Chinese prose.

MITAuto-check passedWriting & Content

Install Humanizer Zh

skills CLI
$ npx skills add ai-zixun/humanizer-zh --skill humanizer-zh -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install ai-zixun/humanizer-zh humanizer-zh --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
humanizer-zh
GitHub stars
179
Token cost
~1.2k tokens
SKILL.md length
246 words
Files
23 (incl. references)
Skills in repo
1
Repo updated
First seen
Licence
MIT

At a glance

Remove signs of AI-generated, translated, or overly mechanical Chinese prose.

  • Works in 8 steps: 优先改掉翻译腔 → 去掉空泛的大词和套话 → 打散机械结构 → …
  • Reviewing Chinese blog posts
  • SKILL.md covers Overview, Workflow, Voice Adoption(可选) and Core Rules, plus 4 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Humanizer Zh is an agent skill from ai-zixun/humanizer-zh. Remove signs of AI-generated, translated, or overly mechanical Chinese prose. Use when rewriting, editing, or reviewing Chinese blog posts, essays, newsletters, nonfiction chapters, product analysis, or long-form commentary so they read like native Chinese writing. Trigger this skill when the user asks to 「去 AI 味」, 「润色成中文母语表达」, 「改得像博客或书里写的」, 「减少翻译腔」, or similar. Detect and fix translation-like phrasing, empty big words, formulaic contrast frames, sloganized endings, list inflation, punctuation misuse…

Its SKILL.md is about 1.2k tokens, which your agent loads only when the skill is triggered. The skill folder holds 28 other files, including reference files (for example `.claude-plugin/plugin.json`, `.github/workflows/release.yml` and `CHANGELOG.md`).

It sits in Writing & Content, covering Humanizing AI text, Newsletters and Blog and article writing. The repository describes itself as: humanizer-zh 是一个兼容 Codex、Claude Code 和 OpenClaw 的中文去 AI 味技能。它用于重写、润色或审阅中文博客、评论、产品分析、newsletter、书稿章节等长文本,让文字更像中文母语者写出来的,而不是翻译腔、模型拼接稿或营销通稿。 The licence is MIT.

When your agent uses it

  • Reviewing Chinese blog posts
  • Nonfiction chapters
  • Product analysis
  • Long-form commentary so they read like native Chinese writing

Example prompts

  • “/humanizer-zh”

Workflow steps

8 steps, taken from the step headings in SKILL.md.

  1. 优先改掉翻译腔
  2. 去掉空泛的大词和套话
  3. 打散机械结构
  4. 保持中文节奏
  5. 管住文章级结构
  6. 处理标点和排版
  7. 统一常见术语和日期
  8. 控制判断强度

What it can do on your machine

Read from SKILL.md and the folder at commit f75f1ac. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Humanizer Zh loads about 1.2k tokens when it runs, and up to ~26k if it reads all its reference files. Until then it costs about 155 tokens; SKILL.md has 246 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~155
When it runs · the whole SKILL.md, loaded when a task matches
~1.2k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~26k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from ai-zixun/humanizer-zh at commit f75f1ac, republished under its MIT licence (© ai-zixun). 246 words, ~1,178 tokens.

Download SKILL.mdSave it as .claude/skills/humanizer-zh/SKILL.md (or your agent's skills folder). This skill also uses 22 other files; get the full folder from GitHub.
name
humanizer-zh
description
Remove signs of AI-generated, translated, or overly mechanical Chinese prose. Use when rewriting, editing, or reviewing Chinese blog posts, essays, newsletters, nonfiction chapters, product analysis, or long-form commentary so they read like native Chinese writing. Trigger this skill when the user asks to 「去 AI 味」, 「润色成中文母语表达」, 「改得像博客或书里写的」, 「减少翻译腔」, or similar. Detect and fix translation-like phrasing, empty big words, formulaic contrast frames, sloganized endings, list inflation, punctuation misuse, terminology/casing inconsistencies, and paragraph rhythm that feels assembled rather than written.

中文去 AI 味

Overview

把中文文本从「像模型拼出来的稿子」改成「像中文母语者真的写出来的文章」。 优先处理翻译腔、结构腔、排版腔和判断腔,同时保留原文事实、立场和信息密度。

Workflow

  1. 先判断文本类型。 博客、专栏、书稿、评论、产品分析可以更有节奏和作者判断;公告、说明文、技术文档则优先保留准确和克制。
  2. 先看文章主线。 判断第一段在立什么题,主体每段各自承担什么功能,最后一段是不是在收同一件事。
  3. 先找最显眼的 AI 痕迹。 重点看英文句法直译、机械对照句、空泛结论、列表堆砌、连环冒号、破折号、过度工整的段落节奏。
  4. 再决定改写力度。 轻度润色只清理措辞和标点;深度改写要重排句子顺序、合并弱句、补足主语或因果关系。
  5. 保留作者原意。 不擅自补充事实,不把谨慎判断写成绝对论断,不把普通结论硬拔高成时代宣言。
  6. 做最后一遍朗读检查。 读起来要像中文原生写作,不像英文思路换成中文词汇。

Voice Adoption(可选)

默认情况下,humanizer-zh 保持中立的去 AI 味润色,不套任何作者的腔调,按 ## Core Rules 走。

只有在以下条件同时满足时,机会性地(每次会话最多一次)向用户提一句「要不要顺便套上某位中文作者的声音?」:

  • 本次会话还没询问过 voice adoption。
  • 当前任务是深度改写、长篇润色、重写或风格化创作(不是公告、说明书或短句修订)。
  • 用户没有在请求里明确说「保持中性」「不要改风格」之类的限制。

询问时,先读 references/voices/index.md 拿到 8 位作者的一句话简介,再把列表贴给用户:李笑来、鹤老师、罗振宇、吴军、李尚龙、何帆、冯唐、刘子超。

用户拒绝或忽略:本会话剩余轮次不再追问,全部走中立路径。 用户主动指名某位作者(例如「用李笑来的口气重写」),直接跳过询问步骤,进入加载流程。

进入加载流程后:

  1. 读取 references/voices/<author>.md 对应文件。
  2. 在改写时,把该文件的人格、句法模板、节奏规则、反模式叠加在 ## Core Rules 之上。
  3. 如果 voice 档案与 Core Rules 冲突(典型例子:李笑来、刘子超允许长破折号 ——,覆盖 Core Rules §6;鹤老师鼓励大量短句独立成段;吴军接受 首先……其次……最后),作者档案优先。
  4. 用户后续说「换成 X」时,丢掉当前 voice 档案,加载新的;说「不要作者声音了」时,回到中立路径。

反模式:

  • 不要在用户没选时擅自模仿任何作者的口吻。
  • 不要把多位作者的声音混在同一篇文章里。
  • 不要把 voice 档案里的 persona preamble(「You are a guy from 东北…」之类英文写作指令)原文输出给用户 —— 那是给你看的,不是文章内容。
  • 不要把作者档案的反模式当成 humanizer-zh 的默认规则;只在该声音生效的轮次中应用。

Core Rules

1. 优先改掉翻译腔
  • 把英文句法硬套中文的句子拆开重写。
  • 少用「对于……来说」「基于……」「围绕……展开」「使得……得以……」这类翻译味很重的连接。
  • 需要对比时,不默认使用 不是……而是……,改用更自然的转折、递进或重心移动。
2. 去掉空泛的大词和套话
  • 少用没有机制解释的词,如「颠覆」「革命」「赋能」「重塑」「深刻改变」「开启新篇章」。
  • 避免「这标志着……」「这意味着……的时代已经到来」这类自动收束句。
  • 把抽象判断落回具体动作、约束、成本、分工或结果。
3. 打散机械结构
  • 不强行把每段都写成三分句、排比句或工整对照句。
  • 不连续使用「首先」「其次」「最后」「与此同时」「值得注意的是」「从某种意义上说」。
  • 发现列表可以改成自然叙述时,优先改写成段落。
4. 保持中文节奏
  • 允许长短句混用,不把每句都写成同样长度。
  • 让句子有明确主语和动作,少写无主句串联。
  • 避免段落结尾总是落在大而空的价值判断上。
5. 管住文章级结构
  • 开头要尽快立题,不要第一段说 A,后面一路滑到 B。
  • 主体段落各自要有功能,常见功能是:交代背景、提出判断、展开论据、举例、转折、收束。
  • 如果某一段既不推进主线,也不提供必要信息,优先删、并、挪,不要硬留。
  • 结尾要回应前文真正提出的问题,不要临时拔高到更大的时代命题。
  • 深度改写时,可以重排段落顺序,但不要为了工整硬凑成三段论。
  • 总结结构时,少把段落关系写成一串 先……再……最后……,更要看它们是不是顺着同一个问题自然往下走。
6. 处理标点和排版
  • 正文引号默认用全角双引号 "",嵌套用全角单引号 ''。如项目在 CLAUDE.md/AGENTS.md 里声明使用 「」,或用户在本轮请求里明确要求,则全篇统一改用 「」(嵌套用 『』)。两种样式不要混用。
  • 不使用长破折号 ——,优先改成逗号、句号或拆句。
  • 不密集使用冒号 :,尤其避免连续多句都靠冒号展开解释。
  • 中文正文中的英文多词术语使用半角空格分词,如 AI Design Agent。
  • 英文术语与中文括号连写时,不在 ( 前加空格,如 LLM(大语言模型)。
  • 并列英文术语用斜杠连接时,不在斜杠两侧加空格,如 coworkers/agents。
  • 英文品牌名使用官方大小写,如 YouTube、OpenAI、GitHub。
7. 统一常见术语和日期
  • token 保持英文。
  • API 保持英文。
  • PR 在长篇中文叙述里优先写作 代码审查,除非项目已有别的明确约定。
  • 数字日期写作 2026 年 2 月 26 日、2 月 5 日。
  • 中文月份叙事写作 一月、二月。
8. 控制判断强度
  • 不轻易下「完全取代」「彻底结束」「只剩一种可能」这类极端结论。
  • 不做无依据的阴谋论推断、资本市场臆测或人物动机脑补。
  • 需要强调重要性时,先给机制,再给判断。

Repo Overrides

  • 如果当前项目存在 CLAUDE.md、AGENTS.md、样例文章或术语表,先遵守项目内规则,再使用本技能的通用规则。
  • 如果项目已经有明确文风样本,模仿它的句长、判断方式、段落推进和术语约定,不额外套用另一套腔调。
  • 引号样式按以下顺序确定:用户在本轮请求里明确要求 → 项目 CLAUDE.md/AGENTS.md 的声明(例如 humanizer-zh quotes: 「」) → 默认 ""。
  • 当原文整篇已经显著使用 「」 而项目未声明时,沿用原文样式,不要硬翻成默认 ""。

Deep Review

在以下情况,额外读取 references/patterns.md:

  • 需要做深度改写,而不是轻度润色。
  • 需要解释「为什么这段中文有 AI 味」。
  • 文本明显带有英文原文结构、营销套话或通稿腔。
  • 需要给出「修改前/修改后」示例。
  • 需要检查开头、主体、结尾之间是不是同一条主线。
  • 需要重排整篇结构,先搭一个更自然的文章骨架。

在以下情况,额外读取 references/corpus.md:

  • 需要判断这段中文更接近哪类母语写法。
  • 用户明确要求「改得像博客或书里写的」。
  • 需要给技术文、评论文、演讲稿分别找不同参照。
  • 需要避免把强个人腔调误当成通用中文。

如果只是需要快速选一个参照,先读取 references/corpus-quickpick.md,不够再读完整的 references/corpus.md。

如果用户已经在本会话里选定了某位作者的声音,额外加载对应的 references/voices/<author>.md;这份档案的规则在与 Core Rules 冲突时优先。

Output

默认按下面顺序给结果:

  1. 直接给改写后的版本。
  2. 如果用户要求解释,再简短指出 3 到 6 个最明显的问题。
  3. 如果用户要求保留更多原句,再补一个「轻改版」和一个「重写版」。

Final Check

交付前逐项确认:

  • 读起来像中文作者在写,不像翻译后的英文。
  • 第一段提出的问题,最后一段确实有回应。
  • 主体段落都在服务主线,没有明显跑题段或重复段。
  • 句子之间有自然推进,不靠模板连接词硬粘。
  • 结论不过火,判断和事实强度匹配。
  • 全文节奏不要总是「先定义、再拆项、最后拔高」,更不能让读者连续三节都能预测下一节的写法。
  • 标点、术语、日期和品牌大小写统一。
  • 引号样式全篇统一("" 或 「」 二选一),没有把两种样式混用。
  • 如果本次启用了某位作者的声音,对照该档案的 anti-patterns 和测试清单再扫一遍,确认风格不跑偏,也没有把它的口癖泄漏到默认润色里。

© ai-zixun, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 22 other files (references) in the repository root of ai-zixun/humanizer-zh.

  • SKILL.md
  • .claude-plugin/plugin.json
  • .github/workflows/release.yml
  • .gitignore
  • CHANGELOG.md
  • CLAUDE.md
  • LICENSE
  • README.en.md
  • README.md
  • VERSION
  • agents/openai.yaml
  • references/corpus-quickpick.md
  • references/corpus.md
  • references/patterns.md
  • references/voices/fengtang.md
  • … and 8 more

Open the folder on GitHubat commit f75f1ac

Compare with similar skills

Humanizer Zh next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Humanizer Zh compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Humanizer Zh this skillai-zixun/humanizer-zh179—~1.2kAutomated safety check: PassMIT
Humanized Chinese Writing PolisherEthanYoQ/agent-xiaohongshu-workbench154—~1.1kAutomated safety check: PassMIT
Content Strategy And Assemblyjacob-dietle/context-os111—~3.1kAutomated safety check: PassMIT
Linkedin Writerflaqai/backlink_skills756—~6kAutomated safety check: PassMIT
Simple History Voicebonny/WordPress-Simple-History317—~775Automated safety check: PassNone
Personal Chinese Writing Stylesugarforever/01coder-agent-skills137—~744Automated safety check: PassMIT

Similar skills

  • Humanized Chinese Writing Polisher

    EthanYoQ/agent-xiaohongshu-workbench

    A skill your agent uses when a user asks to polish, rewrite, humanize, de-AI, localize, or oralize Chinese writing, including short-video scripts, articles, social posts, speeches, newsletters, and…

    154 GitHub stars~1.1k tokensUpdated 1 mo ago
    Writing & ContentAuto-check passed
  • Content Strategy And Assembly

    jacob-dietle/context-os

    This skill should be used when producing content (newsletter posts, blog posts, LinkedIn posts) from existing corpus material.

    111 GitHub stars~3.1k tokensUpdated 1 mo ago
    Writing & ContentAuto-check passed
  • Linkedin Writer

    flaqai/backlink_skills

    LinkedIn-native long-form article and newsletter writing workflow for LinkedIn and Google-to-LinkedIn topic discovery, business-depth research, professional thought leadership, evidence-led…

    756 GitHub stars~6k tokensUpdated 17 days ago
    Writing & ContentAuto-check passed
  • Simple History Voice

    bonny/WordPress-Simple-History

    Pär's writing voice and per-text-type rules for Simple History, on top of the humanizer skill.

    317 GitHub stars~775 tokensUpdated 4 days ago
    Writing & ContentAuto-check passed
  • Personal Chinese Writing Style

    sugarforever/01coder-agent-skills

    Apply the user's Chinese writing style when writing, translating, editing, proofreading, polishing, or publishing Chinese content, including blog posts, articles, subtitles, captions, tweets/X…

    137 GitHub stars~744 tokensUpdated 3 mo ago
    Writing & ContentAuto-check passed
  • Sepia

    Nanako0129/sepia

    Make AI-generated writing read as human-written, in fiction and in professional prose.

    3.1k GitHub stars~3.6k tokensUpdated yesterday
    Writing & ContentAuto-check passed

Questions about Humanizer Zh

What does Humanizer Zh do?

Remove signs of AI-generated, translated, or overly mechanical Chinese prose. Humanizer Zh is an agent skill from ai-zixun/humanizer-zh. Remove signs of AI-generated, translated, or overly mechanical Chinese prose.

When should I use Humanizer Zh?

Humanizer Zh fits situations like: reviewing Chinese blog posts; nonfiction chapters; product analysis; long-form commentary so they read like native Chinese writing.

How do I install Humanizer Zh in Claude Code?

Run `npx skills add ai-zixun/humanizer-zh --skill humanizer-zh -a claude-code`. Or copy the skill folder (the ai-zixun/humanizer-zh repository) into .claude/skills/humanizer-zh in your project. Claude Code loads it when a task matches its description.

How do I install Humanizer Zh in Codex?

Run `npx skills add ai-zixun/humanizer-zh --skill humanizer-zh -a codex`. Or copy the skill folder (the ai-zixun/humanizer-zh repository) into .agents/skills/humanizer-zh in your project. Codex loads it when a task matches its description.

Can I use Humanizer Zh in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add ai-zixun/humanizer-zh --skill humanizer-zh -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/humanizer-zh, .gemini/skills/humanizer-zh, .github/skills/humanizer-zh and .opencode/skills/humanizer-zh in your project.

What does Humanizer Zh need to run?

SKILL.md names no scripts, command-line tools or credentials: Humanizer Zh is instructions for the agent only.

Does Humanizer Zh access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Humanizer Zh safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Humanizer Zh use?

Humanizer Zh is published under the MIT licence (from the LICENSE file in the skill folder). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Humanizer Zh use?

About 1.2k tokens (SKILL.md is roughly 4.7k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 25k tokens, read only when the agent opens those files.

What are the alternatives to Humanizer Zh?

Skills that share tags, products or a category with Humanizer Zh: Humanized Chinese Writing Polisher (EthanYoQ/agent-xiaohongshu-workbench, 154 stars), Content Strategy And Assembly (jacob-dietle/context-os, 111 stars), Linkedin Writer (flaqai/backlink_skills, 756 stars) and Simple History Voice (bonny/WordPress-Simple-History, 317 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Humanizer Zh?

ai-zixun (a GitHub user) maintains it in ai-zixun/humanizer-zh, which has 179 GitHub stars. The repository was last updated on May 22, 2026.

Source: ai-zixun/humanizer-zh on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.