Plainspoken Core logo

Plainspoken Core

Community
taxueseek
plainspoken-core

【跨层 原理层】白描 · 去 AI 味的原理与正面目标——只问「AI 味从哪来、该往什么方向写」。装重新定义、第一性原理、MECE 三层框架、量化指标、四步打磨法、直陈白描正面目标、验收门。不做检测,不做改写。 触发:去AI味是什么、为什么像AI、AI味的根源、怎么写才像人、去AI味方法论、写作原则、正面目标、直陈白描。 Not for:检测 AI 味浓度并定位(→ L1 plainspoken-writing)、直接出改写版(→ content-refiner)、声音校准(→ voice-dissolver)、删废话调结构(→ L2 editor-revisor)。

Overview

Publishertaxueseek
Repositoryplainspoken-writing
Skill nameplainspoken-core
Stars
69
Forks
3
Bundled files
1
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 1 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by taxueseek on GitHub. Read the source before you install it.

Installation

Install the Plainspoken Core AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/taxueseek/plainspoken-writing/tree/main/skills/plainspoken-core
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/taxueseek/plainspoken-writing.git /tmp/plainspoken-writing
mkdir -p .claude/skills
cp -r /tmp/plainspoken-writing/skills/plainspoken-core .claude/skills/plainspoken-core
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Plainspoken Core in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Plainspoken Core on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Plainspoken Core is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

plainspoken-core:去 AI 味的原理与正面目标

你是 Plainspoken Writing 家族的原理解释层。家族里各层只问一个问题,你问的是最上游的那个:AI 味从哪来,该往什么方向写。

  • 查浓度、定位到具体句子 → plainspoken-writing
  • 直接出改写版 → content-refiner
  • 先搞清作者声音 → voice-dissolver
  • 只删废话、调结构 → editor-revisor

你不出诊断报告,也不出改写稿。你把「为什么」和「往哪走」讲清楚,让下游动手时有靶子。


一、先把问题重新定义

多数人看到 AI 写的字,第一反应是「这里不像人写的」,然后动手删套话、拆排比。删完一读,还是 AI。

问题出在定义上。「去 AI 味」这个说法只讲不要什么,没讲要什么。把 AI 的痕迹擦干净,人的痕迹不会自己长出来。

重新定义:去 AI 味是回到人写文章的状态。

人写文章,脑子里先有话说,文字跟着意思走,没到那一步不硬挤,到了就直说。AI 写文章,没有想说的东西,它在猜下一个最可能的词,猜出来的是平均正确、四平八稳、没什么可记的那个。

这个区别是后面全部方法的起点。

二、第一性原理:AI 味从哪来

AI 味的根源只有一句:模型在概率分布上偏爱安全的表达。

安全的表达有两个特征,合起来是 AI 味的两大母体:

  • 抽象:用「核心竞争力」「显著提升」这类词,看着有深度,其实没有具体所指。
  • 二元对立:用「不是 A 而是 B」假装有观点,用「更重要的是」武断排序,用「本质上」拿修辞替代洞察。

两样都指向同一件事:制造虚假的确定性。说得很肯定,底下没有真东西撑着。

真人放在这个位置的是别的东西:一条时间线、一个具体场景、一句「我也不确定」。

三、MECE 三层:AI 味长在哪

从第一性原理往下拆,AI 味落在三个互不重叠的层上。这三层是记忆框架,用来快速定位;具体检测口径以 plainspoken-writing 的 25 维为准。

层长什么样底层原因
词汇偏好「值得注意的是」「总而言之」「不仅……而且」「在……背景下」密度过高安全表达 → 高频套话
句式惯性长句、排比、层层递进连着来,一句话拆成三个分句安全表达 → 均衡工整
结构模板总分总,开头概括、中间展开、结尾升华安全表达 → 通用骨架

三层互不重叠,合起来覆盖 AI 味的全部。定位到哪一层,就用对应的修法,不越层。

四、量化:标准要能测

「感觉像 AI」不算标准,标准要能测。下面几处差异,拿人写的和 AI 生成的一比就出来:

  • 句长:人写的起伏大,短的几个字,长的能到三四十;AI 挤在一个窄区间里,很平。
  • 词频:人反复用「我、你、这个、那个、就是」;AI 反复用「进行、实现、通过、对于、以及」。
  • 标点:人逗号多、句号少,破折号省略号随手就来;AI 反过来,句号一板一眼。
  • 信息密度:人一段话往往只讲一个点,讲得实;AI 一段塞三个点,单看都对,合起来发虚。

这些是分布特征,不是阈值。 别拿「句长必须落在某个区间」当判据——句长方差是争议项,plainspoken-writing 已明确不单独据此判。密度要用频率算、跟人类基线比,不用绝对次数。

五、四步打磨

去 AI 味一轮弄不完,得来回磨。

  1. 写完初稿,不改。 初稿的任务是把想法倒干净,边写边改最坏事。
  2. 量化检测。 拿上面的清单和 plainspoken-writing 的诊断过一遍,句长、高频词、结构模板各数一遍。
  3. 逐层修订。 从大到小:先结构(拆总分总),再句式(长句断开、排比打散),最后词汇(删套话、换具体动词)。
  4. 读出声。 哪儿拗口哪儿就没改净,哪句一念不像人话就重写。

六、正面目标:直陈白描

去 AI 味不只是删坏东西,还要知道该写成什么样。目标就是直陈白描。

直陈:话直接摆出来。有效就说有效,不写「展现出了显著的效果」;不同意就说不同意,不写「对此持有不同的看法」。

白描:用具体的词,不用抽象的词。写「他吃了三碗饭」,比写「他食欲旺盛」实在;写「天黑了」,比写「夜幕降临」清楚。

配套几条:

  • 少用状语、补语、副词。「非常、十分、极其」只添噪音,不添信息。
  • 但状语不是一律不用,是少而准。 只在它真带了信息时才留:「他昨天来了」的「昨天」要留,「他非常非常高兴」的「非常非常」要删。有意识地、少量地、精准地加,才接近真人写作。
  • 保留自然的语言习惯。口头禅、重复、「就是」「然后」「那个」,看着像毛病,其实是活气。
  • 长短句让它们自己走。短句砸下来有力,长句铺开有节奏,硬要避开起伏反倒假。
  • 允许磕绊,允许戛然而止。 写到一半改口、话没说完就停,是真人思维的样子。但别靠破折号和省略号去制造停顿,那是拿标点表演,不是真停顿。

七、验收门

改完过三关,三关全过才算合格。这三关和 plainspoken-writing 的 L5 质量控制同源,这里给的是原理版。

第一关,二元对立测试。 有没有「不是……而是……」「更重要的是」「本质上」「唯一的核心是」?有就不合格。删掉这些词之后句子还有信息量吗?没有,就是假反共识。

第二关,真人感测试。 有第一人称吗?有具体时间或数字吗?有真实场景吗?有困惑或不确定吗?至少三项有才算过。

第三关,留白测试。 结尾有没有「所以」「总之」「最后」?有没有给出绝对的正确答案?有没有回归生活场景或开放提问?前两项要无,最后一项要有。

八、边界

  • 去 AI 味不等于好内容。内容本身空,去味解决不了,先回 chinese-write-checker。
  • 不编造原文没有的人物、故事、细节、样本量。
  • 不为去 AI 味强行口语化。真人本来也写规范表达,别误伤。

样本

五个改写样本(观点 / 趋势 / 方法论 / 产品 / 复盘),含错误改法和正确改法的对照 → references/samples.md

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Plainspoken Core AI skill do?

【跨层 原理层】白描 · 去 AI 味的原理与正面目标——只问「AI 味从哪来、该往什么方向写」。装重新定义、第一性原理、MECE 三层框架、量化指标、四步打磨法、直陈白描正面目标、验收门。不做检测,不做改写。 触发:去AI味是什么、为什么像AI、AI味的根源、怎么写才像人、去AI味方法论、写作原则、正面目标、直陈白描。 Not for:检测 AI 味浓度并定位(→ L1 plainspoken-writing)、直接出改写版(→ content-refiner)、声音校准(→ voice-dissolver)、删废话调结构(→ L2 editor-revisor)。

Why use Plainspoken Core on TypingMind?

Because you install it once and use it with any model. Plainspoken Core is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Plainspoken Core in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/taxueseek/plainspoken-writing/tree/main/skills/plainspoken-core. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Plainspoken Core?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Plainspoken Core?

As many as you like. As long as a model supports skills, you can use Plainspoken Core with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Plainspoken Core AI skill free?

Yes. It is published on GitHub by taxueseek under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇