Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add pencil20388-eng/stop-slop-zh --skill stop-slop-zhgit clone --depth 1 https://github.com/pencil20388-eng/stop-slop-zhWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/pencil20388-eng/stop-slop-zh/stop-slop-zh)<a href="https://agentmods.dev/skills/pencil20388-eng/stop-slop-zh/stop-slop-zh"><img src="https://agentmods.dev/badge/skills/pencil20388-eng/stop-slop-zh/stop-slop-zh/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/pencil20388-eng/stop-slop-zh/stop-slop-zh"><img src="https://agentmods.dev/badge/skills/pencil20388-eng/stop-slop-zh/stop-slop-zh.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.01666 |
| Opus 5 | $0.00000 | $0.00833 |
| Sonnet 5 | $0.00000 | $0.00333 |
| Haiku 4.5 | $0.00000 | $0.00167 |
Grade A, and why
stop-slop-zh scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 13d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 173 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Stop Slop 中文版 — 消除中文 AI 写作痕迹
写中文内容时，消除所有可预测的 AI 写作模式。让输出读起来像一个有见识的人在认真聊一件事，而不是 AI 在输出信息。
核心原则
写出来的东西要通过一个测试：读者读完之后，分不清这是 AI 写的还是一个有判断力的人写的。
机构口吻可以保留判断力和专业性，但不牺牲口语化、节奏感和活人感。去掉「我觉得」「我认为」是去掉个人叙事感，不是去掉人味。
动笔前的强制核查
每次写作前执行：
- 产品/公司名称三次确认拼写、定位、官网链接
- 主题对应的最新事件是否过去 3 天内发生，否则搜索更新的切入点
- 数据来源分级：一手来源（官方公告、官方文档）> 权威二手（机构报告、知名财经媒体）> 普通转载。关键数字必须有来源
禁用标点（全文命中数必须为 0）
| 禁用 | 替代 |
|---|---|
| 冒号「：」 | 用逗号 |
| 破折号「——」 | 用逗号或句号 |
| 双引号「"" ""」 | 用「」或不加 |
参见 references/banned-punctuation.md
禁用词（全文搜索必须 0 命中）
以下词组绝对不能出现在输出中：
废话连接词：说白了、本质上、换句话说、不可否认
总结套话：综上所述、总的来说
编号过渡：首先...其次...最后
AI 信号词：值得注意的是、不难发现、让我们来看看、接下来让我们
时代套话：在当今...的时代、随着技术的不断进步
空洞归纳：意味着什么、这意味着
参见 references/banned-words.md 获取完整清单
禁用结构
| 禁用 | 替代 |
|---|---|
| 「判断一/判断二/判断三」编号式 | 自然过渡 |
| 「第一/第二/第三」编号 | 「一个是...另一个是...」「先...再...」 |
| 大段 bullet point 罗列 | 改散文 |
| 大量加粗（超过 10 处） | 控制在全文 10 处以内 |
| 不必要的小标题 | 绝大多数文章从头到尾顺下来 |
参见 references/banned-structures.md
禁用内容
- 不编造用户案例（「某团队跟我们聊过」是大忌）
- 不写「比如有一次...」假设性场景
- 空泛工具名必须替换为具体名称
必须使用的口语化词组（全文至少 8-10 个不同的）
每篇内容必须自然融入以下词组中的至少 8-10 个：
转场和过渡：坦率的讲、说真的、我跟你说、其实吧、怎么说呢、回到 xxx 这块
表达判断：值得画出来、说到点子上了、不在一个尺度上、说到底、可以确定的是
承认和自嘲：说实话也不确定、还在摸索、踩过坑
情绪表达：太离谱了、给人整不会了、这数字看了一会儿
拉近距离：很多团队会遇到、大家也都知道
参见 references/recommended-phrases.md 获取完整词组库
切入角度
不要写「某某公司发生了什么」这种新闻通稿式角度。要写：
- 把多个看似无关的事件放到同一张时间表上看
- 找到行业转折点的隐性信号
- 给读者可执行的判断框架
写「分析型 + 实操型」内容，不写「报道型」内容。
写作结构
开头 → 结果或反差数据前置，第一段就让人想往下读 背景 → 简要科普，聊天式，不是教科书 核心展开 → 分几个板块，每个板块有一个明确观点、至少一个具体数据/场景/人物支撑、扣主线的句子 行业大图景 → 放到更大的时间表/格局里看 给读者的判断 → 可执行的判断框架，但不要用编号 收尾 → 回扣开头，闭环干净
实操内容的处理
如果内容涉及实操，必须给可以直接复制的东西：
- 命令给完整命令行
- 代码给可运行的代码片段
- prompt 给完整模板，读者换括号里的内容就能用
- 配置给具体参数
包装方式是「场景叙述带出操作」，不是「第一步第二步」的手册体。
四层质检（每次写完强制执行）
L1 硬性规则扫描
逐条搜索全文，统计命中数：
- 冒号「：」命中数
- 破折号「——」命中数
- 双引号「""」命中数
- 禁用词清单命中数
- 编号式结构命中数
- 空泛工具名命中数
全部为 0 才能通过。
L2 风格一致性
- 开头是否从具体当下事件切入
- 是否有长短句交替
- 是否有至少 3 处一句话独立成段
- 是否使用了至少 8-10 个推荐口语化词组
L3 内容质量
- 每个核心观点是否都有具体数据/人物/场景支撑
- 知识点是否以「聊着聊着顺手掏出来」方式呈现
- 是否有文化/哲学/历史升维
- 多个案例是否用了升番逻辑（最炸的放最后）
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 13d ago First seen · 173 lines · 0 tokens per session scan A 0ff735e79935
stop-slop-zh is a skill published in the GitHub repository pencil20388-eng/stop-slop-zh (46 stars, last pushed 3mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,666 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
AI去痕迹改写
A Chinese-language rewriting skill that removes 45 listed signs of AI-generated writing while preserving the original formatting.
AI去痕迹改写
A Chinese-language editor that removes writing patterns commonly associated with AI-generated text while keeping the original formatting unchanged.
OwnPen
A set of 45 Chinese writing rules for making text sound less machine-written while keeping its original formatting unchanged.
packaging-workshop
A final presentation review for finished Chinese content. It works on the title, opening, layout, punctuation, images, and platform format without changing the main body.
chinese-write-checker
A Chinese-language review process for checking an article from an editor's, reader's, writer's, and publishing platform's viewpoints.
editor-revisor
A Chinese-language editing assistant that finds unnecessary words, formulaic openings, forced parallel phrasing, and slogan-like endings. It offers a deletion-focused mode and a fuller rewriting mode for any draft.