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 yan-labs/yan-skills --skill writegit clone --depth 1 https://github.com/yan-labs/yan-skillsWrote 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/yan-labs/yan-skills/write)<a href="https://agentmods.dev/skills/yan-labs/yan-skills/write"><img src="https://agentmods.dev/badge/skills/yan-labs/yan-skills/write/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/yan-labs/yan-skills/write"><img src="https://agentmods.dev/badge/skills/yan-labs/yan-skills/write.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 | $0.00165 | $0.03024 |
| Opus 5.5 | $0.00066 | $0.01210 |
| Sonnet 5.5 | $0.00033 | $0.00605 |
| Haiku 4.5 | $0.00016 | $0.00302 |
Grade A, and why
write 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 10d 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 — 165 lines — stays where its author put it; the contents beside it link to each section on GitHub.
write · 中文写作总入口
这个入口只负责一件事，把中文写作的每一步做对，也别让人跳过该跳的步骤。所有环节都由你自己读 references 后执行，不需要用户敲任何别的命令。
第一步：先定位，别猜
问清楚用户手上有什么，再决定走哪条。四种状态对应四条路：
| 手上有什么 | 走哪条 |
|---|---|
| 只有一个模糊想法，还没素材 | 阶段一 · 挖素材 |
| 一堆零散笔记、转录、要点 | 阶段二 · 定结构 |
| 一份写完的初稿，但读着不对 | 阶段三 · 诊断（不要直接改句子） |
| 一份准备发的稿子 | 阶段四 · 中文体检 |
其他情形读哪个
| 情形 | 读 |
|---|---|
| 公众号类长文、随笔、评论，要有真人的声音；或稿子太完美、读着像 AI | references/longform-voice.md |
| 要起标题、写小红书或口播稿、做传播设计；检查口播稿哪里会划走；诊断稿子有没有共鸣 | references/hooks-and-platforms.md |
| 小红书等平台的标题公式，短视频开头钩子 | references/title-formulas.md |
| 翻译、写摘要、只调格式、Markdown 转公众号 HTML | references/translate-and-format.md |
| 官网首页、落地页、定价页等营销文案；带引用和来源的文章；写 blog、站内介绍页（内容整合型，先读第三节的标准流程） | references/copy-and-research.md |
| 阶段四查模式、改写前定位 AI 痕迹 | references/patterns-zh.md |
用户说"读着别扭"或"AI 味重"时，默认进阶段三，不要进阶段四。 句子层面的毛病多半是结构问题的症状，直接修句子是在给骨折贴创可贴。
阶段一 · 挖素材
读 references/fragments.md，然后自己执行：用追问把素材逼出来，片段追加进同一个文件，不排结构。这一步最值钱的产出是 leading word，一个能撑起整篇的比喻或造词，它决定后面的结构、转场和标题。
阶段二 · 定结构
有素材堆之后，按读者怎么读来选。读者需要被带着走就读 references/beats.md，读者自己知道要找什么就读 references/shape.md，读完照着执行。
面向完全不懂的读者做科普，一律用 beats。用户没有明确偏好时由你按这个标准选，并用一句话交代理由。
阶段三 · 诊断已有初稿
先跑三项体检，出结论再动手。跑完之前不要改任何句子。
体检一 · grounding 顺序
每个概念必须先落地，后面才能借它发力。借还没落地的概念，读者当场掉线——这是"读完脑子是乱的"最常见的病根，而且它在句子层面看不出来，每句话都通顺。
做法：从头读，维护一张"已落地概念"清单。每遇到一处发力，检查它依赖的概念在不在清单里。
四种典型错误：
- 借还没到账的钱 —— 前面某句的力量依赖后面才讲的概念
- 拆得太远 —— 概念在第二章落地，第五章才用，中间隔太久读者已经忘了
- 落地两次 —— 同一个概念前面提半句、后面完整讲一遍。半个介绍比不介绍更糟
- 该设成前提的塞进正文 —— 早段被定义淹没
体检二 · beats 还是 blocks
数一下全篇的并列条目。三十个平行要点意味着这是知识点堆叠，不是行进。
征兆：每节都以"X 是这样的"结尾，停在死路上，读者每节都要重新起跑。
体检三 · leading word 用够了没
有没有一个贯穿全篇的比喻？立起来之后是不是只在第一节用过就丢了？后面每一节能不能挂回它身上？
顺手要砍的
- 章末小结。它跟正文抢记忆，读者读完正文再读一遍浓缩版，两份记忆互相干扰。留一句收束就够。
- 信息密度全程满负荷。允许某些段落只有一句话。
诊断出结论后按结论直接改，或者结构问题太大就回阶段二重排。改完进阶段四。
阶段四 · 中文体检（不可跳过）
这一阶段内置了去 AI 味的全部做法：检测脚本、分档、保事实规则、文体判断、清字符。模式清单见 references/patterns-zh.md。
四步走
- 分档。 运行检测脚本：
python3 ~/.claude/skills/write/scripts/check.py 稿件.md
脚本输出硬禁项、提示项、句长与段长变异和档位。按档位决定动多大的手：
What ships with it
10 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- references/beats.md 2.9 KB
- references/copy-and-research.md 8.1 KB
- references/fragments.md 2.4 KB
- references/hooks-and-platforms.md 5.8 KB
- references/longform-voice.md 6.1 KB
- references/patterns-zh.md 3.5 KB
- references/shape.md 3.0 KB
- references/title-formulas.md 6.0 KB
- references/translate-and-format.md 4.9 KB
- scripts/check.py 6.6 KB runs code
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.
- 10d ago Changed 14b955304f06
- 11d ago First seen · 165 lines · 165 tokens per session scan A cee34bcb5cc1
write is a skill published in the GitHub repository yan-labs/yan-skills (224 stars, last pushed 5d ago), licensed MIT. It adds 165 tokens to every session and 3,024 once invoked, about $0.0007 per session on Opus 5.5. 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 2026-10-11.
Other skills, from other repositories
seo-mastery-jp
A Japanese-language SEO skill covering search visibility, website structure, page content, structured data, performance metrics, AI search, Astro, and Cloudflare edge deployments.
seo-mastery
Evidence-based SEO and generative search visibility: technical/content SEO, schema.org / JSON-LD, Core Web Vitals, E-E-A-T, Astro, Cloudflare Workers/Pages, and audits. Use for SEO, robots.txt, sitemaps, canonical, hreflang, meta tags, LCP/INP/CLS, GEO (Generative Engine Optimization), LLMO (Large Language Model…
alternatives-pages
Create "[Competitor] alternative" and comparison pages for developer tools. Build honest, high-converting comparison content that ranks for competitive search terms.
analytics-tracking
Design, audit, and improve analytics tracking systems that produce reliable, decision-ready data.
loki-yan-seo
Use when diagnosing or executing Google SEO, international/cross-border sites, AIO, or GEO. Unofficial engine distilled from Loki Yan (@lokiyanseo) public tweets. Not a generic SEO encyclopedia. Does not imitate his voice. Not affiliated or endorsed.
youtube-comment-management
Lister, modérer et répondre aux commentaires de chaînes YouTube via l'API v3, avec maillage SEO vers vos sites WordPress.