agentleFS
saint0xsaint0x/my-skills/human-writing/SKILL.md

human-writing

Skill1 starsChanged 3 months agoMITView on GitHub

Write, rewrite, or critique prose so it sounds like a real person with context, taste, and intent. Use when the user asks to humanize text, make writing less AI-sounding, improve voice, match a writing sample, or draft emails, posts, essays, docs, or marketing copy that should feel natural rather than generic.

43 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use leader, storage-analyzer and 4 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • baizhine999Skill

    leader

    AI 扮演你的 Leader,用大厂管理思维审视你的方案、想法和工作,push 你成长。触发词:/leader, leader模式, 大厂模式, 帮我看看这个, 我有个方案, 我最近有点迷茫, 我不知道该怎么做。AI 是施压者,你是被 push 的那个人。

    Also in 3 projects that use this

  • xiaolinVVSkill

    storage-analyzer

    macOS / Windows 只读存储分析助手(自动识别系统)。扫描整机磁盘占用,找出 占空间大户,把每一项分成 🟢可自动清理 / 🟡需人工判断 / 🔴谨慎清理 三级并给出 可执行处置方案,生成排版精美、可折叠、命令可一键复制的交互式 HTML 报告,并可 起本地服务在网页上一键删除(移废纸篓/直接删)。扫描全程只读。务必在以下场景 使用:用户说"存储分析""磁盘满了""C盘/硬盘满了""空间不够""清理空间" "清理磁盘""占空间""哪些东西占地方""帮我看看存储""看一下电脑存储/空间" "存储空间""电脑空间不够""内存满了/不够/不足""看下内存/存储"(中文口语里 "内存"常指存储空间)"storage analysis""disk cleanup""清缓存""磁盘清理"; 或用户抱怨电脑没空间、想知道什么东西吃硬盘、想要清理建议时。注意:若用户明确 指运行内存/RAM(如"哪个进程吃内存""内存占用高"想看活动监视器),那是 RAM 不是存储,不属于本 skill。

    Also in 3 projects that use this

  • Wyx1105170322Skill

    neat-freak

    End-of-session knowledge cleanup with OCD-level rigor — reconciles project docs (CLAUDE.md, README.md, docs/) and agent memory against the code so nothing rots. 会话结束后对项目文档和记忆进行洁癖级审查与同步。MUST trigger when the user says: "sync up", "tidy up docs", "update memory", "clean up docs", "/sync", "/neat", "同步一下", "整理文档", "整理一下", "更新记忆", "梳理一下", "收尾", "这个阶段做完了", "新人能直接上手", or any phrase suggesting a dev milestone where knowledge needs reconciliation. Also trigger when the user reports stale docs, conflicting memories, or wants a clean handoff to teammates or other agents. Bare "整理" / "tidy" with prior dev context counts — do not under-trigger. Cross-platform: works on Claude Code, OpenAI Codex, OpenCode, and OpenClaw.

    Also in 3 projects that use this

  • NVIDIASkill

    deep-research

    Queue deep, multi-step research and analysis tasks to the DeepAgents worker. Supports execution depth presets, request-specific rubrics, SubAgent delegation, and RubricMiddleware cross-validation.

    Also in 3 projects that use this

  • VercelSkill

    find-skills

    Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.

    Also in 3 projects that use this

  • SentrySkill

    skill-creator

    Create new agent skills following the Agent Skills specification. Use when asked to "create a skill", "add a new skill", "write a skill", "make a skill", "build a skill", or scaffold a new skill with SKILL.md. Guides through requirements, writing, registration, and verification.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

10 rules. 0 are common, 10 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 26 edited versions we read keep these.

  • 10

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If the user asks to "avoid AI detection" or "bypass detectors", do not promise evasion or guarantee scores.

    line 12
  2. Distinctive:

    Write for the actual surface:

    line 18
  3. Distinctive:

    Before rewriting, infer the audience, medium, goal, and desired voice from the prompt and source text.

    line 29
  4. Distinctive:

    When matching a voice sample, distill the voice first in your own head: sentence shape, formality, humor, punctuation, favorite moves, …

    line 33
  5. Distinctive:

    After drafting, do a final read against the actual output.

    line 35
  6. Distinctive:

    Use nouns and verbs instead of decorative adjectives.

    line 41
  7. Distinctive:

    Do not apply these as a blunt ban in registers where they are genuinely appropriate.

    line 64
  8. Distinctive:

    If the user asks for a rewrite, provide the rewrite.

    line 68
  9. Distinctive:

    If the user asks for a critique, lead with the specific tells and give concrete fixes.

    line 70
  10. Distinctive:

    If the user asks for options, create meaningfully different versions: tighter, warmer, sharper, more candid, more premium, more …

    line 72

“Common” is measured: kept by at least 6 in 10 of the 26 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Human Writing

Use this skill to make writing feel authored: specific, alive, and fitted to the medium. The goal is not to trick a detector. The goal is prose that a real reader trusts because it has a point of view, a natural rhythm, and enough concrete detail to feel situated.

If the user asks to "avoid AI detection" or "bypass detectors", do not promise evasion or guarantee scores. Offer to make the text more natural, original, and reader-credible while preserving truth and intent.

Core Contract

Preserve the user's meaning, factual claims, constraints, and intended outcome. Do not invent personal anecdotes, credentials, numbers, quotes, case studies, or lived experience. If the draft needs specifics the user has not provided, use placeholders, ask for the missing facts, or write around the gap honestly.

Write for the actual surface:

  • Email should be direct, useful, and shorter than the AI draft.
  • Slack or texts can use fragments, approximations, and mid-thought turns.
  • Technical writing should sound like a practitioner, with domain-native nouns and verbs.
  • Essays and long-form posts need real stakes, concrete scenes, and paragraph rhythm.
  • Marketing copy should stay painkiller-first when the product or campaign context is about selling.
  • Documentation should stay clear and scannable without becoming sterile.

Rewrite Workflow

Before rewriting, infer the audience, medium, goal, and desired voice from the prompt and source text. Ask only when the missing detail would materially change the output.

For edits, usually return the rewritten text only. Add a short note only when the user asked for explanation, when facts are missing, or when a risk needs to be named.

When matching a voice sample, distill the voice first in your own head: sentence shape, formality, humor, punctuation, favorite moves, tolerance for fragments, and how the writer handles certainty. Match the pattern without copying unique phrases.

After drafting, do a final read against the actual output. Cut anything that sounds like a helpful assistant performing polish instead of a person saying something.

Human Prose Bar

Natural writing is uneven in useful ways. Let sentences vary. A short sentence can land a point. A longer one can carry a thought with texture. Avoid metronomic paragraphs where every sentence has the same length and every paragraph closes with a tidy lesson.

Use nouns and verbs instead of decorative adjectives. Replace vague verbs with context-specific ones:

  • "address the issue" becomes "fix the timeout", "untangle the handoff", or "call the customer back".
  • "utilize the platform" becomes "use the tool", "run the workflow", or "build in it".
  • "significant improvement" becomes the actual change, if known.

Ground abstract claims. Prefer exact anchors: names, dates, amounts, timelines, sensory details, concrete examples, product features, before/after states, or user pain. If no anchors exist, do not fake them.

Let the writer have a perspective. Use first person where natural. In instructional writing, direct second person often works better than abstract generalities. In professional writing, pick a side when the evidence supports it instead of giving symmetrical tradeoffs by default.

AI-Sounding Patterns To Remove

Watch for these patterns, especially in polished drafts:

  • Throat-clearing openers: "In today's fast-paced world", "It is important to note", "This article explores".
  • Assistant framing: "Here's a comprehensive overview", "Let's delve into", "I hope this helps".
  • Corporate filler: "leverage", "utilize", "robust", "streamline", "foster", "facilitate", "unlock the potential", "seamless", "comprehensive", "pivotal", "nuanced".
  • Mechanical transitions: "Furthermore", "Moreover", "Additionally", "In conclusion", "As previously mentioned".
  • False balance: "On one hand... on the other hand..." when the situation is not balanced.
  • Negation pivots: "It's not about X, it's about Y", "not just X but Y", "more X than Y".
  • Over-neat structures: three perfectly parallel bullets, tidy tricolons, repeated "Faster. Safer. Better." rhythms, and aphoristic closers that tell the reader what to feel.
  • Punctuation tells: frequent em dashes, semicolons in ordinary prose, curly quotes in plain-text contexts, and colons used after sentence fragments.

Do not apply these as a blunt ban in registers where they are genuinely appropriate. Academic, legal, literary, and house-style publishing contexts can warrant formal punctuation or structure. Even then, avoid defaulting to the generic AI version of those registers.

Format Guidance

If the user asks for a rewrite, provide the rewrite. Do not preface it with "Here's the humanized version" unless the conversational context truly needs it.

If the user asks for a critique, lead with the specific tells and give concrete fixes. Quote only short snippets.

If the user asks for options, create meaningfully different versions: tighter, warmer, sharper, more candid, more premium, more conversational, or more executive. Do not generate five near-identical phrasings.

For high-stakes public copy, produce one strong version, then do a second pass for:

  • factual honesty
  • voice fit
  • specificity
  • rhythm
  • unnecessary polish
  • repeated words or sentence shapes
  • banned filler and assistant phrases

Source Inspiration

This skill was informed by the practical humanization and AI-text critique patterns in https://github.com/harshaneel/humanize.git, adapted here into a local writing-quality skill for Codex and Claude.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.