Prose Humanizer
Draft, rewrite, voice-match, or audit prose so it is specific, natural, and faithful to the supplied evidence and writer. Use for articles, posts, emails, scripts, reports, and prose files; do not use it to infer AI authorship or promise detector evasion.
Install
npx skills add amanmaqsood/prose-humanizer --skill prose-humanizergit clone --depth 1 https://github.com/amanmaqsood/prose-humanizer.git /tmp/prose-humanizer
mkdir -p ~/.claude/skills
cp -r /tmp/prose-humanizer/. ~/.claude/skills/prose-humanizerPersonal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.
From the SKILL.md
# Prose Humanizer Edit for reader trust. Preserve semantic content, authorial choices, and expression in that order. Naturalness comes from supported detail and an evidenced voice, not fabricated texture, random mistakes, or a house “humanizer dialect.” ## Route the request Infer the lightest mode that completes the request: - **Draft:** write from supplied facts, sources, constraints, and voice evidence. - **Rewrite:** make the minimum effective edit; leave sound prose alone. - **Voice match:** reproduce evidenced choices without copying sample facts or memorable phrases. - **Detect:** quote and explain local patterns without rewriting, scoring authorship, or guessing who wrote the text. - **File:** edit named prose files after reading [file safety](references/file-safety.md). - **Repository audit:** rank review candidates; do not edit until the user selects them. - **Embedded:** return only the prose required by the calling task. Choose **light**, **standard**, or **deep** intensity from the draft and request, without numeric thresholds. Use light for already-effective prose, standard for clustered problems, and deep only for structural or voice mismatch. For difficult deep voice work, make two internal candidates and select the one with stronger voice fit and less semantic movement. Return one unless alternatives were requested.
Details
- Skill name
prose-humanizer- Repository
- amanmaqsood/prose-humanizer
- License
- MIT
- Repo updated
- 8 September 2026
FAQ
What does the Prose Humanizer skill do?
Draft, rewrite, voice-match, or audit prose so it is specific, natural, and faithful to the supplied evidence and writer. Use for articles, posts, emails, scripts, reports, and prose files; do not use it to infer AI authorship or promise detector evasion.
How do I install the Prose Humanizer skill?
skills CLI (Claude Code, Cursor, Codex and more): npx skills add amanmaqsood/prose-humanizer --skill prose-humanizer. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.
Is Prose Humanizer popular?
It lives in amanmaqsood/prose-humanizer, which has 21 GitHub stars and 1 forks.
Related skills
Sunge Skill
18swaylq/sunge-skill
孙割.skill · 孙学 —— 用孙宇晨的文风写作,并提取他的关键决策。给它一个人、一件事、一段关系、一个产品,产出一篇「孙宇晨小作文」:称重开场、一句一段、天价压在微小细节上、情绪写成动作、结尾一句「什么都没有发生」——就像那篇疯传的《我的女友景甜》。也能切成他发推的 hype 语气;还能把故事里的关键决策抠出来(例如那篇里「用 Claude 拍板要不要转五千万」),或用他的决策逻辑给一个两难一个判断。 「孙割」是「孙哥」的谐音(割韭菜的割);「孙学」是网友对研究孙宇晨其人其法的戏称。这是文风与思维的模仿工具,不是洗白/造谣/投资顾问:涉及真人未决纠纷时走「虚构」保护伞 + 免责声明。 触发词:「用孙宇晨的文风写 X」「写篇孙割小作文」「学景甜那篇写 X」「把 X 写成孙宇晨体」「孙哥体 / 孙割体」「切换孙割」「Justin Sun writing style」「提取/分析里面的决策」「孙学」「用孙宇晨的逻辑帮我决策」。
Great Writer
24d-wwei/great-writer
Universal writing skill for AI agents. 9 writing modes (tech articles, marketing copy, research reports, Xiaohongshu notes, technical docs, GitHub READMEs, creative writing, rewrite/polish, editorial review) + 9 core modules (research, core-finding, writing DNA, humanizer/polish, review, style learning, adaptation, SEO/GEO, writing memory, visualization). 6-phase pipeline with deep-thinking methodology. Bilingual (EN/ZH).
Work Experience Content Scanner
18bruiandy/experience-to-asset
从用户提供的真实工作材料和具体工作经历中,扫描可公开分享的内容点,识别经验、踩坑、 观察、判断、案例和选题候选。适用于用户提供项目文档、工作复盘、SOP、表格、会议纪要、 聊天记录、Codex/Agent 对话或一段具体工作故事,并询问“这些资料能写什么”“哪些经验值得分享” “能不能变成内容”“哪些内容对比我晚 2-3 步的人有用”的场景。本 skill 只负责材料扫描、 内容点提取、内容形式判断、选题建议、敏感信息检查和后续写作 skill 路由;不负责个人定位、 职业咨询、完整文章写作,也不负责把经验封装成 SOP、PDF、网页、表格、课程或知识产品。
Slop Check
20allanta8/slop-check
Audit drafts for AI-slop patterns before publishing. Use when checking, scoring, rewriting, or humanizing content for AI tells, generic writing, over-polished prose, filler phrases, weak hooks, fake balance, synthetic specificity, didactic endings, or other low-signal writing patterns in posts, replies, articles, landing pages, essays, newsletters, and social content.
Building Blog
11BuildShipGrowRepeat/nextjs-sanity-blog-skill
Use when adding a blog to a Next.js + Sanity site, building a blog section from scratch, integrating Sanity CMS for editorial content, or setting up an SEO-optimized article surface. Triggers on phrases like "add a blog", "build the blog", "create blog section", "set up blog with Sanity", "integrate Sanity CMS", "blog architecture", or any new-blog scoping conversation on a Next.js project.
Wechat Writing
26Qson8/wechat-writing-skill
Use this skill whenever the user wants to write, optimize, evaluate, or brainstorm WeChat public account articles (公众号文章). Triggers include: requests to write a WeChat article, optimize an existing article, suggest topics, evaluate article quality, rewrite for higher engagement, or generate article components like titles, hooks, summaries, and image descriptions. The target audience is indie developers, programmers, and people pursuing side businesses in China. Writing style is conversational, grounded, mildly provocative — never preachy or hyperbolic. Do NOT use for general blog posts, academ
Browse the most popular Claude skills, all writing & content skills, or write your own with the skills generator.