anti-vibe-writing
Make AI writing sound human, in English and 中文. A Claude Code / Codex / any-agent skill that strips templated phrasing, consultant-speak, markdown clutter, em-dashes, and copy-paste model residue while keeping the meaning.
Safety scan
A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.
Badge for your README
Markdown
[](/listings/weijt606--anti-vibe-writing)HTML
<a href="/listings/weijt606--anti-vibe-writing"><img src="/api/listings/weijt606--anti-vibe-writing/badge.svg" alt="findsafeskills safety rating" /></a>Install
Install instructions from the README
As written by the repo's author. findsafeskills has not verified these steps.
At its core this is just a set of markdown rules (`SKILL.md` + `references/` + `assets/`), not tied to any one agent. Claude Code, Codex, Kimi, work-buddy, Hermes. If it can read files or take an instruction prompt, it can use this.
**Claude Code, one step:**
/plugin install anti-vibe-writing --marketplace weijt606/anti-vibe-writing
Then call `/anti-vibe-writing`, or just ask Claude to de-AI a draft. On Claude Code older than 2.1.275, run `/plugin marketplace add weijt606/anti-vibe-writing` first, then `/plugin install anti-vibe-writing@anti-vibe-writing`.
**Codex, Cursor, Gemini CLI and other agents, one step with the GitHub CLI** (gh 2.90 or newer):
gh skill install weijt606/anti-vibe-writing anti-vibe-writing --agent codex --scope user
Swap `codex` for `cursor`, `gemini-cli`, `claude-code`, `opencode` or any other name in `gh skill install --help`. It installs the latest release, and `gh skill update` pulls newer ones later. No GitHub CLI? Follow the three steps below.
**1. Get the rules**
Clone the repo, or just copy the `skills/anti-vibe-writing/` folder:
git clone https://github.com/weijt606/anti-vibe-writing.git
**2. Feed it to your agent (pick whichever fits)**
- **Simplest: just hand it the repo link.** No need to clone first: send `https://github.com/weijt606/anti-vibe-writing` and let the agent read `SKILL.md` and `references/` and configure itself. Works with any agent that can browse the web or run git. - **One-off (any agent)**: open `skills/anti-vibe-writing/assets/rewrite-prompt-template.md`. It has ready-made instruction blocks: Full Rewrite / Light Cleanup in English, and "中文改写(带负向约束)" for Chinese. Copy the block for your language and send it with your draft. - **Persistent**: put `SKILL.md` and the matching `references/*patterns-to-remove.md` wherever your agent loads context. Names differ by tool: - Agents with a skills directory (e.g. Claude Code): drop it in `~/.claude/skills/anti-vibe-writing/`, then call `/anti-vibe-writing` - Agents with a project-instructions file (e.g. Codex's `AGENTS.md`): write or include the rules there - Otherwise: paste into the system prompt / custom instructions / knowledge base
**3. (Optional) Name the scenario and mode**
One line of context changes the result a lot: - Scenario: "this is a tweet / a newsletter / a technical memo" - Loosen up: "make it feel personal" / "blog voice" → enables human-texture mode - Match a voice: paste a few of your own samples and say "learn my style" → enables learning mode
When in doubt, say nothing. The default clean mode is right for most drafts.
What was scanned
The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.
.claude-plugin/marketplace.json · Manifest · 16 lines
- {
- "name": "anti-vibe-writing",
- "owner": { "name": "weijt606", "url": "https://github.com/weijt606" },
- "metadata": {
- "description": "Make AI writing sound genuinely idiomatic, in English and Chinese."
- },
- "plugins": [
- {
- "name": "anti-vibe-writing",
- "source": "./",
- "description": "Bilingual (English + 中文) skill that removes the AI-generated feel from drafts while preserving meaning.",
- "version": "1.9.0",
- "category": "writing"
- }
- ]
- }
README.md · README · 196 lines
- # anti-vibe-writing
- 
- [](./README.md)
- [](./README.zh.md)
- [](./LICENSE)
- [](./CHANGELOG.md)
- > 本 README 提供中英文两个版本:**English**(本页,默认)和 **[中文](./README.zh.md)**。点上方徽章切换。
- > **One goal: make AI sound genuinely idiomatic, 倍儿地道.**
- An agent writing skill that removes the AI-generated feel from documents. Works in English and Chinese, with optional modes for matching a specific author's voice.
- 
- <sub>A real pair from [`examples/03-tweet.md`](./examples/03-tweet.md). More in [examples/](./examples/).</sub>
- It runs best as a final pass after drafting with Claude Code, Codex, or any LLM-backed agent. Its job is to keep the substance and remove the tells: templated phrasing, vague abstraction, consultant-speak, markdown-heavy formatting, over-structured outlines, the copy-paste residue chat and search models leave behind (`oaicite` / `turn0search0` / `[cite: 1]` / stray `【】` markers), and the cautious over-balancing that makes writing feel assembled.
- If the default "clean" mode is not enough, the skill also supports:
- - **Human-texture mode**: inject controlled irregularities (inversions, particles, half-sentences, non-standard punctuation) for personal voice
- - **Learning mode**: build a reusable host profile from real samples so future drafts sound like the person who would write them
- - **Scenario presets**: format and tone constraints tuned for tweets, Weibo, blogs, podcast show notes, and professional reports
- ## Quick start
- At its core this is just a set of markdown rules (`SKILL.md` + `references/` + `assets/`), not tied to any one agent. Claude Code, Codex, Kimi, work-buddy, Hermes. If it can read files or take an instruction prompt, it can use this.
- **Claude Code, one step:**
- ```text
- /plugin install anti-vibe-writing --marketplace weijt606/anti-vibe-writing
- ```
- Then call `/anti-vibe-writing`, or just ask Claude to de-AI a draft. On Claude Code older than 2.1.275, run `/plugin marketplace add weijt606/anti-vibe-writing` first, then `/plugin install anti-vibe-writing@anti-vibe-writing`.
- **Codex, Cursor, Gemini CLI and other agents, one step with the GitHub CLI** (gh 2.90 or newer):
- ```bash
- gh skill install weijt606/anti-vibe-writing anti-vibe-writing --agent codex --scope user
- ```
- Swap `codex` for `cursor`, `gemini-cli`, `claude-code`, `opencode` or any other name in `gh skill install --help`. It installs the latest release, and `gh skill update` pulls newer ones later. No GitHub CLI? Follow the three steps below.
- **1. Get the rules**
- Clone the repo, or just copy the `skills/anti-vibe-writing/` folder:
- ```bash
- git clone https://github.com/weijt606/anti-vibe-writing.git
- ```
- **2. Feed it to your agent (pick whichever fits)**
- - **Simplest: just hand it the repo link.** No need to clone first: send `https://github.com/weijt606/anti-vibe-writing` and let the agent read `SKILL.md` and `references/` and configure itself. Works with any agent that can browse the web or run git.
- - **One-off (any agent)**: open `skills/anti-vibe-writing/assets/rewrite-prompt-template.md`. It has ready-made instruction blocks: Full Rewrite / Light Cleanup in English, and "中文改写(带负向约束)" for Chinese. Copy the block for your language and send it with your draft.
- - **Persistent**: put `SKILL.md` and the matching `references/*patterns-to-remove.md` wherever your agent loads context. Names differ by tool:
- - Agents with a skills directory (e.g. Claude Code): drop it in `~/.claude/skills/anti-vibe-writing/`, then call `/anti-vibe-writing`
- - Agents with a project-instructions file (e.g. Codex's `AGENTS.md`): write or include the rules there
- - Otherwise: paste into the system prompt / custom instructions / knowledge base
- **3. (Optional) Name the scenario and mode**
- One line of context changes the result a lot:
- - Scenario: "this is a tweet / a newsletter / a technical memo"
- - Loosen up: "make it feel personal" / "blog voice" → enables human-texture mode
- - Match a voice: paste a few of your own samples and say "learn my style" → enables learning mode
- When in doubt, say nothing. The default clean mode is right for most drafts.
- ## Quick examples
- Curated before/after snippets live in [examples/](./examples/) for anyone scanning the repo. The full regression set stays under `references/`.
- ## Voice modes
- The skill runs in one of three modes:
- | Mode | When to use | Triggered by |
- |---|---|---|
- | Default (clean) | Most product, docs, and professional copy | Default, no opt-in needed |
- | Human texture | Personal blogs, founder notes, social posts | "Loosen it up" / "blog voice" / "make it feel personal" |
- | Learning mode | Series content, personal newsletters, voice-consistent comms | User provides samples, or asks "learn my style" |
- Modes combine with scenario presets (tweet / Weibo / blog / podcast / report). Conflict resolution rules are documented in `SKILL.md`.
- Human-texture mode also has a **social-only "casual typing" (随手打) layer** (off by default; turn it on by naming it, "casual typing" / "开随手打", or describing the effect, "like a quick phone post"; matched by intent, not a fixed phrase, and a plain "loosen it up" won't trigger it): a tiny amount of phone-typing texture on casual posts (dropped end punctuation, no capitalization, an omitted particle) that **never touches numbers, names, or links** and never makes meaning-changing typos. It's phone-typing texture, not error injection to dodge AI detectors.
- ## Use cases
- **The flagship case: Chinese posts on X.** Chinese AI-smell is most obvious on X: 赋能 / 打通, 首先 / 其次, three-clause parallelism, and machine-translation syntax give it away at a glance. This skill is built for exactly that: take an "obviously AI-written" Chinese post and make it read like something a person actually typed. See [`examples/07-tweet-zh.md`](./examples/07-tweet-zh.md) and [`examples/08-translationese-zh.md`](./examples/08-translationese-zh.md).
- Other common cases:
- - Social posts (X, Weibo, Jike, RedNote)
- - Blogs, newsletters, public WeChat articles
- - Podcast show notes and video scripts
- - README cleanup
- - Product docs and landing page copy
- - Proposals, founder notes, technical memos, internal reports
- ## Output goals
- - More human rhythm
- - More intentional structure
- - Cleaner phrasing
- - Stronger voice
- - Less AI smell
- - Sounds like a person chose every word
- ## Repository layout
- ```text
- .claude-plugin/
- plugin.json # Claude Code plugin manifest
- marketplace.json # lets `/plugin install` find it
- agents/
- README.md
- anti-vibe-writing-dev.agent.md # local, gitignored
- anti-vibe-writing-dev.agent.example.md
- skills/
- anti-vibe-writing/
- SKILL.md
- references/
- patterns-to-remove.md # English AI-smell
- chinese-patterns-to-remove.md # 中文 AI 味
- before-after-benchmarks.md # English benchmarks
- chinese-before-after.md # 中文基准
- common-problems-and-fixes.md
- human-passes.md
- human-texture.md # Optional irregularity
- learning-mode.md # Sample-driven style learning
- scenario-presets.md # Per-scenario constraints
- assets/
- final-pass-checklist.md
- rewrite-prompt-template.md
- host-profile-template.md # Fillable host profile
- style-extraction-prompt.md # One-shot extraction prompt
- examples/
- ...
- scripts/
- check-own-prose.sh # CI: project prose passes its own rules
- CHANGELOG.md
- CONTRIBUTING.md
- README.md # English (default)
- README.zh.md # Chinese
- LICENSE
- ```
- ## Working with the files
- Skill files are versioned. The developer agent file is local and gitignored by default, so contributors can adjust it without changing the public repository.
- To create a local agent file, copy `agents/anti-vibe-writing-dev.agent.example.md` to `agents/anti-vibe-writing-dev.agent.md`.
- If you want tool-specific auto-discovery, you may still need to mirror these files into the locations required by the target agent platform.
- ## Credits & references
- This skill stands on the shoulders of several open de-AI / humanizer projects and writeups. The patterns below were studied and adapted into this skill's own structure; the original analysis and wording belong to their authors. Thanks to:
- English:
- - [blader/humanizer](https://github.com/blader/humanizer): a 30-pattern humanizer skill (MIT). Informed the sentence-level tells: copula avoidance, negative parallelism, synonym cycling, false ranges, signposting, diff-anchored writing.
- - [hardikpandya/stop-slop](https://github.com/hardikpandya/stop-slop): AI-slop detection skill (MIT). Source of the optional five-dimension scoring pass (Directness / Rhythm / Trust / Authenticity / Density) in `assets/final-pass-checklist.md`.
- 中文 / Chinese:
- - [op7418/Humanizer-zh](https://github.com/op7418/Humanizer-zh): a 24-pattern Chinese humanizer skill, itself a Chinese adaptation of blader/humanizer (MIT). Informed the copula-"是" avoidance and synonym-cycling tells in the Chinese track.
- - ["AI 中文翻译腔" by yage.ai](https://yage.ai/share/ai-chinese-translationese-20260418.html): the analysis of Chinese translationese (物理动作动词写抽象 / 形容词加冒号预判读者 / 抽象名词主语). Informed the 翻译腔层 of `references/chinese-patterns-to-remove.md`.
- - [@dotey on X](https://x.com/dotey/status/2022774029220749538): discussion of de-AI prompt techniques (role-setting, negative constraints) that shaped the 改写心态 section and the Chinese rewrite prompt block.
- AI-writing research (added in 1.9.0):
- - [Wikipedia: Signs of AI writing](https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing): the community-maintained catalog of AI-writing signs (October 2026 revision). Source of the mid-2026 vocabulary band, source-gap hedging, trailing participle analysis, placeholder and `utm_source` residue, the "Challenges and Future Outlook" ending, and the weak-evidence list.
- - [Graphite's 2026 model-tell analysis, reported by TechCrunch](https://techcrunch.com/2026/10/01/opus-5-5-loves-to-tell-you-this-matters-and-other-ai-writing-tells/): per-model frequencies behind "this matters" / "why X matters", "dependable", the 2026 contrast-frame variants, and the em-dash decline.
- - [《消除"罪证":给写作去除"AI味"的不完全手册(2026版)》](https://eu.36kr.com/zh/p/3824601267196037) by 蒋雨恬, first published on the WeChat account 有三思 U Sense and reprinted by 36kr: informed 过度拔高, the "最" (豆包体) rule, 学术套话 openers, 虚假范围, and the note about people misjudged over two dashes.
- These are independent projects and writeups with their own scope; this repo borrows their ideas and none of their code. If you maintain one of them and want a credit adjusted, open an issue.
- ## Contributing
- See [CONTRIBUTING.md](./CONTRIBUTING.md) for how to add new patterns, benchmarks, or scenario presets.
- ## License
- This project is open source under the MIT License. See `LICENSE`.
- ## Contributor notes
- - Preserve meaning. Sharpen the writing without changing facts.
- - Prefer concrete edits over generic style advice.
- - Keep structure only when it helps the reader.
- ## What's new
- **1.9.0**: catches up with how models write in late 2026 ("this matters", "not simply X", fewer em-dashes, Chinese 过度拔高 and "最" overuse, placeholders and tracking links), adds a list of features too weak to call a text AI-written on their own, and fixes places where the skill broke its own rules. Each version's full notes are in [Releases](https://github.com/weijt606/anti-vibe-writing/releases), and the whole history is in [CHANGELOG.md](./CHANGELOG.md).
Ratings and comments
Sign in to rate this skill.
Sign inFree to use. If it’s helped, pass it on.