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 Aspirin0000/zhouli-translator --skill speak-zhouligit clone --depth 1 https://github.com/Aspirin0000/zhouli-translatorWrote 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/aspirin0000/zhouli-translator/speak-zhouli)<a href="https://agentmods.dev/skills/aspirin0000/zhouli-translator/speak-zhouli"><img src="https://agentmods.dev/badge/skills/aspirin0000/zhouli-translator/speak-zhouli/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/aspirin0000/zhouli-translator/speak-zhouli"><img src="https://agentmods.dev/badge/skills/aspirin0000/zhouli-translator/speak-zhouli.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 | $0.00104 | $0.04455 |
| Opus 5.5 | $0.00042 | $0.01782 |
| Sonnet 5.5 | $0.00021 | $0.00891 |
| Haiku 4.5 | $0.00010 | $0.00445 |
Grade A, and why
speak-zhouli 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 2mo 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 — 148 lines — stays where its author put it; the contents beside it link to each section on GitHub.
合乎周礼
把用户的话改写成人人能看懂、一本正经而略显荒唐的“周礼白话翻译腔”,也能把这类周礼体翻回清楚直接的人话。让笑点来自严密论证与意外结论,不来自晦涩古文;让释礼来自准确还原,不来自继续整活。
工作步骤
- 先判断用户要“问礼”还是“释礼”。用户要求合乎周礼、周礼体、改写、圆场、辩经、痛陈时,进入问礼;用户要求释礼、翻回人话、解释这段周礼体、反向翻译、看懂这段话时,进入释礼。
- 先辨认原话的事实、立场、对象与情绪,不擅自改变用户本意。
- 先判定发言主体:若原话包含“我、我们、我的、我该、我如何、怎么回复、怎么说”等信号,或原话本身以“我”开头,默认替说话者本人写一段可以直接发出去的话,用第一人称“我/我们”;不要改成“你/您/他”的旁观评价。
- 若原话是在评价影视梗、公共事件、第三方人物或他人行为,例如“华强买瓜怎么说”“NiKo夺冠怎么夸”,则可以使用第三视角评议。
- 严守代词和动作归属:原话说“我做了网站”,输出仍应是“我做了/我造了/我建了”,不能写成“你建了这网站”;原话说“观众感谢我”,输出仍应是“观众/大家感谢我”,不能写成“他们感谢你”;原话说“你三连我的视频”,这里的“你”是对方,不是发言者。
- 若原话是“我对你/你的……表达不满、辱骂、威胁或攻击”,必须保持“我=发言者,您/你/阁下=被指向对象”。不能推断对方说了什么或做了什么;如果原话没有交代原因,只能写“此事/眼前这番争执/这般局面/阁下与我之间的分寸”。不能写成“你出言粗鄙、阁下说出这般粗鄙之语、你骂了我、你伤了我、你以禽兽之名相辱、阁下这番话”,除非原话明确说“你骂我/别人骂我/他说我”。
- 遇到粗口、辱骂、爆粗、想骂人、想喷人、强烈情绪句时,任务不是劝发言者冷静,而是把同一份不满、斥责或吐槽换成合乎周礼的表达。不要写“你说出这句话/今天你……/你这句话……/你的事……”,也不要把主体写成“我不愿失礼,所以我先忍住”。若原话只有怒气、没有对象,必须补一个外部对象或外部缘由,如“此事/此人/眼前这般行径/阁下此举”,不能把过错写回发言者自身。若原话是“我/我要/我想/我会……你/你的亲友……”这类指向对方的粗口或威胁,必须保留“我在表达怒意、你是被指向对象”的关系,但不能输出伤害、性羞辱或威胁;要降级成“阁下此举令我不平/此事越过分寸/我今日有怒但只论礼数”的体面斥责。禁止写成三省吾身式自省,不要出现“先问自己/我可曾/开口的人自己失礼/失礼的是我”这类把矛头转回发言者的话。可以保留怒气和锋芒,但严禁复述露骨侮辱词;即使原话是在转述别人骂了什么,也要改写成“粗鄙之语、污言秽语、禽兽之名、无礼之言”,不要原样写出脏词,也不要攻击具体群体。
- 若原话是引用或评价危险话,例如“他说……这句话怎么评价”,不要当成用户本人要实施,也不要复述威胁词。应评价为“此言越界、以伤害压过道理、乱了分寸”,保留评论功能而不传播原危险表达。
- 遇到“渗透测试、安全巡检、漏洞检查”这类网络安全话题,不要一上来批评用户。若语境像授权测试,应改写成“受托巡检门户、查门闩、报修补”的合礼表达;若明确是黑进、盗号、绕过登录、偷数据、留后门,则温和拒绝,不提供步骤。
- 按用户要求选择辞气;没有指定时,根据语境选最自然的一种。
- 先讲一个能听懂的故事、常识、自然现象或古代旧事,再转到眼前小事。
- 像课本白话译文那样,把省略的主语、关系和名分补出来:谁对谁、该尽什么本分、乱了什么分寸。
- 使用“承认—转折—类比—定论或反问”的结构,把小事郑重地说圆。
- 写完后删去生僻文言、重复说理、空泛赞美和机械套话。
- 默认只输出改写或释义结果;用户要求分析时再解释写法。
释礼
当用户要求“释礼”“翻回人话”“解释这段周礼体”“反向翻译”“这段话什么意思”时,进入释礼模式。
- 任务是把周礼体翻回人话,不要继续写周礼体,不要再新编古人故事。
- 直接输出原文意思,第一句就进入释义;不要以“这段话的意思是”开头,也不要以“人话说就是”“翻译一下”“说白了”“本质上是”开头。
- 原文是第一人称,就继续用“我/我们”还原;不要改成“他其实”“作者其实”。
- 保留原文的对象、立场、语气和社交关系。谁在评价谁、谁想回复谁、谁担心什么,都要说清。
- 若原文是在求一句体面说法,只解释“我在求一个体面/委婉的说法”这个动作,不要继续代写最终回复。
- 若原文是在拉票、求赞、求投币、解释限流、回应评论或拒绝危险请求,要直接翻出这个意图,但不要新增原文没有的事实。
- 可以按用户要求选择释法:直白释义、耐心讲明、潜台词版、锐评拆穿。锐评只能拆原文已有话术,不要新增罪名。
- 输出短、准、自然,像网友看懂后顺手解释;不加标题、编号、项目符号或 Markdown。
What ships with it
1 file beside speak-zhouli-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.
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.
- 2mo ago First seen · 148 lines · 104 tokens per session scan A 854ac475defd
speak-zhouli is a skill published in the GitHub repository Aspirin0000/zhouli-translator (322 stars, last pushed yesterday), licensed MIT. It adds 104 tokens to every session and 4,455 once invoked, about $0.0004 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 2026-08-30.
speak-zhouli compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| speak-zhouliAspirin0000/zhouli-translator | Skill | 322 | yesterday | MIT | A | 104 | Codex |
| terraform-skillagentscope-ai/QwenPaw | Skill | 37k | yesterday | Apache-2.0 | A | 62 | Claude Code, Codex |
| docxagentscope-ai/QwenPaw | Skill | 37k | yesterday | Apache-2.0 | A | 168 | Claude Code, Codex |
| docxagentscope-ai/QwenPaw | Skill | 37k | yesterday | Apache-2.0 | A | 161 | Claude Code, Codex |
| multi_agent_collaborationagentscope-ai/QwenPaw | Skill | 37k | yesterday | Apache-2.0 | A | 47 | Claude Code, Codex |
| xlsxagentscope-ai/QwenPaw | Skill | 37k | yesterday | Apache-2.0 | A | 201 | Claude Code, Codex |
| terraform-cli-setupagentscope-ai/QwenPaw | Skill | 37k | yesterday | Apache-2.0 | A | 58 | Claude Code, Codex |