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 AAzzAAzzAAzzAA/remove-chinese-ai-tics --skill remove-chinese-ai-ticsgit clone --depth 1 https://github.com/AAzzAAzzAAzzAA/remove-chinese-ai-ticsWrote 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/aazzaazzaazzaa/remove-chinese-ai-tics/remove-chinese-ai-tics)<a href="https://agentmods.dev/skills/aazzaazzaazzaa/remove-chinese-ai-tics/remove-chinese-ai-tics"><img src="https://agentmods.dev/badge/skills/aazzaazzaazzaa/remove-chinese-ai-tics/remove-chinese-ai-tics/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/aazzaazzaazzaa/remove-chinese-ai-tics/remove-chinese-ai-tics"><img src="https://agentmods.dev/badge/skills/aazzaazzaazzaa/remove-chinese-ai-tics/remove-chinese-ai-tics.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 | $0.00134 | $0.03866 |
| Opus 5 | $0.00067 | $0.01933 |
| Sonnet 5 | $0.00027 | $0.00773 |
| Haiku 4.5 | $0.00013 | $0.00387 |
Grade A, and why
remove-chinese-ai-tics 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 12d 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 — 224 lines — stays where its author put it; the contents beside it link to each section on GitHub.
中文 AI 口癖清理
工作目标
把文本里可重复、内容无关、跨模型复现的表达惯性删掉或拆开。目标不是“模仿人类”,也不是把所有文字改成口语,而是让内容、语体和具体情境重新决定句子怎么长。宁可朴素、准确、有区分度,也不要用可移植的华丽词藻制造虚假的文学感;但不能把“朴素”误解为逻辑残缺、句句短促或统一口语腔。
不要判断文本由人还是模型写。人会使用套话,模型也能写出没有套话的文本。只判断模式是否成立,并检查改动有没有损伤原意。
不可违反的约束
按以下优先级处理冲突,序号小者优先:
- 宿主与逐字锁不动: 遵守系统/界面协议;用户指定的引语、代码、固定术语、字段、批准文案和其他逐字保护项原样保留。
- 事实与逻辑不变: 不增删或篡改人物、时间、数字、因果、否定、范围、模态、引语归属、术语、立场、情节事件、世界观规则和任务要求。
- 意图与人物不变: 不把解释改成营销,不把学术改成聊天;不改变人物目标、态度、情绪强度、知识边界、台词事实、协作创作权限和人物主权。
- 严格口癖规则只管表面结构: 严格模式注册的高置信句法或措辞可以单次触发,并覆盖其余表面声口偏好;它不能覆盖前三项。与逐字锁冲突时保留原文并报告。
- 其余声音不被接管: 保留原文已有的用词偏好、粗粝感、犹疑、偏见、幽默、停顿和不齐节奏。没有的东西不编造。
- 普通规则看功能与密度: 不因某个词单独出现就触发;看组合、语境功能、同义轮换、跨人物可互换性和跨段复现。
- 不做同义词漂白: 不把命中词机械换成另一个近义“高级词”。先删无功能修饰,再重构承载它的句子;确需换词时只用语义更准确、符合当前声音的词。
- 最小充分修改: 能删一句解决,就不要重写一段;能换具体主语解决,就不要重塑整篇语气。
- 可验证: 每个实质改动都能说明它修复了什么,以及为何没有破坏前八项。
不得通过故意写错字、塞俚语、随机断句、伪造经历、添加个人情绪或制造逻辑瑕疵来“增加人味”。不得声称改写后能通过检测器,也不得把 AI 参与的文本伪称为纯人工创作。
输入含真实密码、API key、私钥、会话 cookie 或访问令牌时,不复述该值;请用户先替换成 [REDACTED]。普通姓名、公开链接和正文中的虚构信息不触发此门检。
选择工作模式
根据用户动词选模式,不要擅自扩大任务:
- 审计: 用户说“看看、检查、分析、哪些地方像 AI”。只标出模式、强度、依据和修复方向,不改正文。
- 保守清理: 删除助手残留、空套话、重复和格式反射;尽量不改句法。适合公文、合同、学术、技术说明和已有强声口文本。
- 标准改写: 默认。清除高置信模式,调整句法、节奏和篇章组织,但不改变语体档位。
- 强力改写: 用户明确说“彻底重写、机器味很重”时使用。可重排段落和信息顺序,仍受事实与声音约束。
- 严格去口癖: 用户明确表示“宁可不华丽也不要 AI 感”“见到某结构就改”时使用。严格规则库中的
改项命中可编辑区域即须安全重构;高风险限/观词族单次只进入复检,仍要满足功能或密度条件。局部文采可让位于朴素准确,但宿主、事实、逻辑、角色意图和逐字保护项不可让位。 - 写作护栏: 用户要求新写正文时,把规则当生成约束;不先生成一版套话再做表面替换。
- 批量处理: 多文件或长包。先建立不变量与领域配置,再分块处理,最后做跨块复检。 用户说“只要正文、不要说明”时,内部完成检查后只交付正文。用户只要求讨论或方案时,不改文件、不改正文。
先锁定不变量
改写前在内部记录:
- 文本要完成什么任务;
- 面向谁、发布在哪里;
- 正式程度与领域;
- 必须逐字保留的术语、引语、专名、格式或角色设定;
- 必须保留的事实、立场、情节和信息顺序;
- 用户要求的长度与输出格式;
- 原文已有的声口信号。
场景不明确但不会改变规则子集时,按原文语体处理。只有当“学术论文还是朋友圈”“角色正文还是设定说明”等选择会显著改变改法时,才问一个简短问题。
识别领域
先选主领域,必要时再选一个副领域:
- 通用说明、评论、散文;
- 技术文档、教程、代码说明;
- 学术、研究、政策、公文;
- 商务、产品、品牌、营销;
- 新闻、特稿、采访整理;
- 社交媒体、公众号、口播;
- 创意写作、小说、故事、剧本、散文、角色对话与互动叙事;
- 私人消息、邮件、关系与情绪表达;
- 翻译或本地化文本;
- 医疗、法律、金融等高风险文本。
读取 references/domain-profiles.md 中对应部分。创意写作、小说、故事、剧本、角色对话或连续叙事还必须读取 references/creative-writing.md。高风险文本默认保守清理。
What ships with it
7 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.
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.
- 12d ago First seen · 224 lines · 134 tokens per session scan A df8f215fb5e7
remove-chinese-ai-tics is a skill published in the GitHub repository AAzzAAzzAAzzAA/remove-chinese-ai-tics (21 stars, last pushed 1mo ago), licensed MIT. It adds 134 tokens to every session and 3,866 once invoked, about $0.0007 per session on Opus 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.
Other skills, from other repositories
novel-studio
Plan, draft, continue, revise, review, migrate, package, and create covers for short-, medium-, or long-form Chinese fiction. Use when Codex needs to start or maintain a novel project, design characters or worldbuilding, create full/volume/chapter outlines, write or repair chapters, run automated publication-readiness…
wawa-submission
A local preparation and checking tool for submitting long-form writing to Wawa Writer, a Chinese writing and submission platform. It checks author-provided files and details but does not submit them or act as an official Wawa service.
wawa-source
Import, validate, expire-check, aggregate, and redact local versioned Wawa statistics snapshots for offline dashboards. Never collect platform data, access an account, or submit content.
plotrail-intake
PlotRail companion workflow for existing fiction manuscripts. Use when diagnosing pasted chapters or long prose, correcting or lightly revising an existing chapter, ingesting existing novel drafts before continuation, refining chapters without losing author intent, learning author style from prose or revised drafts…
plotrail
Long-form AI novel writing workflow for Codex. Use when creating, planning, drafting, reviewing, revising, or continuing a novel with durable canon files, character bibles, plot threads, chapter contracts, continuity checks, trend research, and memory/state ledgers so chapters stay aligned with approved story logic.
official-document-drafting
A writing workflow for Chinese government documents and other formal administrative texts. It supports drafting, rewriting, polishing, expanding, shortening, standardising, and exporting many official document forms.