Edit English and Chinese prose to reduce formulaic AI writing patterns, including filler, inflated claims, and vague attribution. Preserves the original meaning and intended tone while improving sentence variety, directness, and the writer’s voice.
This skill should be used when the user asks to "remove AI writing patterns", "humanize this text", "make this sound more natural", "remove AI-generated traces", "fix robotic writing", or needs to eliminate AI writing patterns from prose. Supports both English and Chinese text. Based on Wikipedia's "Signs of AI writing" guide, detects and fixes inflated symbolism, promotional language, superficial -ing analyses, vague attributions, AI vocabulary, negative parallelisms, and excessive conjunctive phrases.
Run score Not tested
See how to installReview the instructions the agent will receive before you install.
Remove AI-generated writing patterns from text to make it sound natural and human-written. Supports both English and Chinese.
When editing academic prose, check the active installed
skills/ml-paper-writing/references/knowledge/paper-miner-writing-memory.md
under the current client's skill home for relevant source-attributed writing
examples. Use it only to guide natural structure and rhythm. Keep the author's
meaning, evidence, and target venue constraints intact. If the memory is absent
or irrelevant, continue with this skill's own pattern references.
This skill identifies and eliminates predictable AI writing patterns from prose, based on Wikipedia: Signs of AI writing, maintained by WikiProject AI Cleanup.
Core insight: LLMs use statistical algorithms to predict what should come next. The result tends toward the most statistically likely outcome that applies to the widest variety of cases—creating detectable patterns.
Trigger phrases:
Use cases:
Remove throat-clearing openers and emphasis crutches.
English examples:
中文示例:
Avoid binary contrasts, dramatic fragmentation, rhetorical setups.
Patterns to avoid:
Mix sentence lengths. End paragraphs differently.
Check:
State facts directly. Skip softening, justification, hand-holding.
Bad: "It could potentially be argued that the policy might have some effect." Good: "The policy may affect outcomes."
If it sounds like a pull-quote, rewrite it.
Bad: "This represents a major step in the right direction." Good: "The company plans to open two more locations."
| Pattern | Description | 中文描述 |
|---|---|---|
| Undue emphasis | "stands as a testament", "crucial role" | "作为……的证明"，"关键作用" |
| Promotional language | "vibrant", "rich heritage", "breathtaking" | "充满活力的"，"丰富遗产"，"令人叹为观止" |
| Vague attributions | "Experts believe", "Observers note" | "专家认为"，"观察者指出" |
| Superficial -ing analyses | "highlighting the importance", "ensuring that" | "强调……的重要性"，"确保……" |
| Formulaic "challenges" sections | "Despite X, faces challenges" | "尽管……面临挑战" |
| Pattern | Description | 中文描述 |
|---|---|---|
| AI vocabulary | Additionally, crucial, delve, enhance, landscape | 此外，至关重要，深入探讨，增强，格局 |
| Copula avoidance | "serves as", "stands for", "represents" | "作为"，"代表"，"充当" |
| Em dash overuse | Using — more than humans | 过度使用破折号 |
| Rule of three | Forcing ideas into groups of three | 强行三段式 |
| Elegant variation | Excessive synonym substitution | 过度换词 |
For comprehensive pattern lists, see:
references/patterns-english.md - Complete English pattern referencereferences/patterns-chinese.md - Complete Chinese pattern referenceAvoiding AI patterns is only half the job. Sterile, voiceless writing is just as obvious.
Have opinions. Don't just report facts—react to them.
"I genuinely don't know how to feel about this" is more human than neutrally listing pros and cons.
Vary your rhythm. Short punchy sentences. Then longer ones that take their time.
Acknowledge complexity. Real humans have mixed feelings.
"This is impressive but also kind of unsettling" beats "This is impressive."
Use "I" when it fits. First person isn't unprofessional—it's honest.
"I keep coming back to..." signals a real person thinking.
中文示例：
"我真的不知道该怎么看待这件事"比中立地列出利弊更有人味。
"这令人印象深刻但也有点不安"胜过"这令人印象深刻"。
Rate the text 1-10 on each dimension (总分 50):
| Dimension | Question | 问题 | Score |
|---|---|---|---|
| Directness | Direct statements or announcements? | 直接陈述还是绕圈宣告？ | /10 |
| Rhythm | Varied or metronomic? | 节奏变化还是机械重复？ | /10 |
| Trust | Respects reader intelligence? | 尊重读者智慧吗？ | /10 |
| Authenticity | Sounds human? | 听起来像真人吗？ | /10 |
| Density | Anything cuttable? | 有可删减的内容吗？ | /10 |
Standard:
See examples/ for before/after transformations:
examples/english.md - English text examplesexamples/chinese.md - Chinese text examples| Before | After |
|---|---|
| "serves as a testament to" | "shows" |
| "Moreover, it provides" | "It adds" |
| "It's not just X, it's Y" | "X does Y" |
| "Industry experts believe" | "According to [specific source]" |
| 改写前 | 改写后 |
|---|---|
| "作为……的证明" | "表明" |
| "此外，……提供了" | "……增加了" |
| "这不仅仅是……而是……" | "……是……" |
| "专家认为" | "根据[具体来源]" |
references/patterns-english.md - Complete English pattern referencereferences/patterns-chinese.md - 完整中文模式参考references/phrases-to-cut.md - Filler phrases to remove (需删除的填充短语)references/wikipedia-source.md - Original Wikipedia source materialexamples/english.md - English before/after examplesexamples/chinese.md - 中文改写示例✅ DO:
❌ DON'T:
MIT
Based on Wikipedia: Signs of AI writing, maintained by WikiProject AI Cleanup. Merges content from humanizer, humanizer-zh, and stop-slop skills.
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Create and configure Claude Code agents using Markdown files with YAML configuration headers. Define when each agent should run, write its responsibilities and workflow, choose models and tools, and check configuration and triggering behavior.
Add datasets, models, and other components to machine learning projects using consistent registration and creation patterns. The skill guides file placement, configuration, and automatic module loading so new components fit the project's established architecture.
Investigate code errors through a structured process using reproduction steps, logs, and testable explanations. The workflow guides fault isolation, fixes, and follow-up checks, with common error patterns for Python, JavaScript/TypeScript, and Bash/Zsh.
Check academic references for existence, accurate publication details, and support for the claims being cited. This guide explains how to verify sources, obtain bibliography entries, and flag references that cannot be confirmed.
Review code changes and pull requests for correctness, security, performance, and test quality. The skill helps prioritize issues, write specific and constructive feedback, and create shared review checklists and standards for a development team.
Create reusable slash commands, prompts invoked by typing a slash, for Claude Code. The skill explains Markdown structure, arguments, file references, tool permissions, and plugin integration so recurring tasks can follow a defined workflow.