基于维基百科 35 种 AI 写作特征Pattern,对文本进行两轮改写以消除 AI 痕迹,保留原文事实不虚构,支持自定义文风样例。
Humanizer 将 AI 风格的文字重新改写,使其读起来像人写的,但不会改变原意。因为它只是 Markdown 格式,所以可以配合任何支持 skills 的 Agent 使用。
Humanizer 使用了 Wikipedia「AI 写作迹象」词条中的 35 种模式,由 WikiProject AI Cleanup 维护。它先不做固定结构假设地通读一遍原文,然后对照这些模式检查草稿,确保在改写任何仍然听起来不自然的内容之前,保留了原文的主张。
"LLM 使用统计算法来猜测下一个应该出现什么。结果倾向于在最广泛情形下最可能出现的答案。"
它不会编造内容。名称、数字、日期、引文、引用或其他事实细节必须来自原文或写作者本人。对于私人写作,Humanizer 会保留作者本人的风格。技术类和参考资料类文章则保持中立和简洁。如果你提供了写作范例,Humanizer 会遵循该范例而非默认风格规则。
粘贴文字后,Humanizer 会在给出最终版本之前展示其推理过程。你会看到第一轮改写以及对仍显生硬之处的简短点评。指向一个文件时,它只会修改正文部分,代码、数据、frontmatter 和链接目标保持不变。
直接调用 skill:
/humanizer
[paste your text here]
或者用自然语言请求:
Please humanize this text: [your text]
要重写文件,给出文件路径:
Humanize the prose in docs/launch-post.md
如果你希望改写结果更像你自己的语气,可以附上范例:
/humanizer
Here's a sample of my writing for voice matching:
[paste 2-3 paragraphs of your own writing]
Now humanize this text:
[paste AI text to humanize]
Humanizer 会遵循范例中的节奏、措辞、标点和刻意的小习惯。
语言和语法模式
月名、街区名等细节需要由写作者本人提供。如果缺少这些信息,Humanizer 应该主动询问,而不是自行编造。
改前(AI 腔):
I recently spent five unforgettable days in Lisbon, and let me tell you — this city completely stole my heart. From the moment I arrived, I knew I was somewhere truly special.
Nestled along the banks of the Tagus River, Lisbon stands as a vibrant testament to Portugal's enduring spirit, where rich history and modern energy intertwine at every turn. Yes, the famous hills are challenging — my legs certainly felt it! — but every climb rewards you with breathtaking, panoramic views that make it all worthwhile.
No trip would be complete without riding the iconic Tram 28, winding through the city's most historic neighborhoods. And the food? Simply divine. The original pastéis de nata at Pastéis de Belém are a beloved national treasure, and savoring one still warm was a moment I will never forget.
But what truly makes Lisbon special isn't just the sights — it's the feeling. Wander a few steps off the main squares and you'll discover a quieter, more authentic side: sun-drenched alleys, charming tiled facades, and friendly locals going about their daily lives, inviting you to slow down and savor every moment.
Of course, the must-see São Jorge Castle offers stunning views over the rooftops below. The queues can be long, but the experience is absolutely worth it for any history buff or curious traveler.
Would I go back? Absolutely. Lisbon isn't just a place to visit — it's a place to fall in love with, again and again. If you're dreaming of your next getaway, this is one destination that promises memories to last a lifetime. ✨
改后(人写腔):
I spent five days in Lisbon last October and still have mixed feelings about it. Beautiful, yes. Also harder on the knees than anyone warned me.
The hills are the whole story and somehow never make the brochures. My hotel was up in Alfama, which photographs beautifully and translates, in practice, to climbing what felt like a six-story staircase every time I wanted coffee. By the second day my calves had opinions. I started planning each walk around which way was downhill, which is a strange way to see a city but probably an honest one.
Everyone says to ride Tram 28, so I did, wedged against a stranger's backpack for forty minutes while three tour groups filmed the same corner. I would walk the route next time, or go before breakfast. The custard tarts, though, earn the fuss. I had one at a plain little place in Graça, still warm, and for about thirty seconds I understood why people build trips around pastry.
What I did not expect was how quiet the city gets away from the main squares. Two blocks uphill from any plaza it turns into laundry lines, chipped tile, open windows, and old men watching football with the sound turned up. That is the Lisbon I keep thinking about, not the castle.
The castle is fine. The view is great, the queue is long, and I spent more time shuffling toward the entrance than looking at anything once I got inside. If I had only two days, I would trade it for an afternoon of getting lost.
I would go back, but in spring and with better shoes. Lisbon does not bend over backward to make things easy for you. I think I liked that, even when my legs disagreed.
Wikipedia: Signs of AI writing 是模式列表的来源。
WikiProject AI Cleanup 负责维护该页面。
2.11.2 - 移除了插件符号链接和独立的 Claude Desktop 包。当前的 Claude Code 直接加载根目录的 SKILL.md,所以 GitHub 的源码 ZIP 现已在 Claude Desktop 中正常工作。35 种模式不变。
2.11.1 - 新增了 Claude Desktop 可用的发布包,内含一个常规的 humanizer/SKILL.md 文件。GitHub 的源码包仍然保留了插件符号链接(修复 #224)。35 种模式不变。
2.11.0 - 用 Plain Language 重写了所有仓库指引、描述、检查和 skill 说明。保留了全部 35 种模式及其行为。
2.10.2 - 为 Claude Desktop 和旧版加载器添加了标准 skills/humanizer/ 插件路径。该路径指向根 skill,所以仍只有一份提示词(修复 #202)。
2.10.1 - 为 §7 添加了 gate、gated、gating 的比喻用法。保留了真实的技术用法,如 feature gating 和 CI quality gates。
2.10.0 - 新增了第 34 和 35 条模式,用于检测残留在终稿中的旧版草稿想法。为真实的限制、反对意见和替代方案添加了保护措施(修复 #198)。同时改进了 §24 和最终改写步骤。共 35 种模式。
2.9.2 - 为第 11 条模式添加了重复句首检测,并对刻意重复添加了保护措施(修复 #206)。扩展了 §28 以覆盖随意公告类语句。共 33 种模式。
2.9.1 - 改进了安装和包检查。移除了不支持的元数据、工具批准和一条重复的长示例。共 33 种模式。
2.9.0 - 新增了禁止编造事实的规则,并更新了所有示例以遵循该规则(修复 #187)。信息重要性高于段落形态,允许写作范例覆盖 §14,并添加了三种输出模式。共 33 种模式。
2.8.3 - 将版本号移至 metadata.version 以兼容 Agent Skills。共 33 种模式。
2.8.2 - 用第一人称里斯本故事替换了主要示例,保留了原主题、视角和细节。共 33 种模式。
2.8.1 - 添加了跨 Agent 安装、Claude 插件文件和引用文字保护措施。共 33 种模式。
2.8.0 - 新增了第 31-33 条模式,并将第 20 条模式扩展以捕获聊天机器人的推销话术。共 33 种模式。
2.7.0 - 新增了第 30 条模式,加强了破折号规则,并将第 21 条模式扩展以覆盖无根据的猜测。共 30 种模式。
2.6.0 - 合并了重复的工作流文本,将人格指引限制在正确的内容类型中,移除了模型猜测,并缩短了主要示例。共 29 种模式。
2.5.1 - 添加了被动语态和缺失主语检测。共 29 种模式。
2.5.0 - 添加了深度真相声称、公告、重复标题和截断的负面结尾。收紧了破折号规则并更正了 frontmatter。共 28 种模式。
2.4.0 - 添加了写作范例匹配功能。
2.3.0 - 添加了连字符词组。
2.2.0 - 添加了草稿检查和第二轮改写。
2.1.1 - 修正了弯引号示例。
2.1.0 - 为全部 24 条模式添加了改前/改后示例。
2.0.0 - 从 Wikipedia 源码重写了 skill。
1.0.0 - 首次发布。
使用 Skills CLI 安装 Humanizer:
npx skills add blader/humanizer --global
省略 --global 则只在当前项目中安装 Humanizer。添加 --agent <name> 或 --agent '*' 可选择接收该 skill 的 Agent,然后重新加载它们的 skills。
Claude Code 2.1.142 或更新版本可直接安装插件:
/plugin marketplace add blader/humanizer
/plugin install humanizer@humanizer
插件命令是 /humanizer:humanizer。
在 Claude Desktop 中,将本仓库下载为 ZIP 并作为 skill 上传。
手动安装:将 SKILL.md 复制到 Agent 的 skill 文件夹中。