Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/Xiaoyangy/novel-studionpx agentmods add skills/xiaoyangy/novel-studio/reviewWrote 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/xiaoyangy/novel-studio/review)<a href="https://agentmods.dev/skills/xiaoyangy/novel-studio/review"><img src="https://agentmods.dev/badge/skills/xiaoyangy/novel-studio/review/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/xiaoyangy/novel-studio/review"><img src="https://agentmods.dev/badge/skills/xiaoyangy/novel-studio/review.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 | $0.00242 | $0.08578 |
| Opus 5 | $0.00121 | $0.04289 |
| Sonnet 5 | $0.00048 | $0.01716 |
| Haiku 4.5 | $0.00024 | $0.00858 |
Grade A, and why
ai-novel-audit 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.
This is a copy
88% identical to smart-search — 347 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 296 lines — stays where its author put it; the contents beside it link to each section on GitHub.
小说 AI 创作度审核
帮助评估一段叙事性文本(小说、故事、散文片段、网文章节等)有多大可能由 AI 生成,或 AI 参与了多少创作。产出一份带证据的概率性评估报告,而非一句「是/否」的判决。
先读这一段:关于可靠性(非常重要)
任何 AI 文本检测都不可靠,本技能也不例外。 在写报告和与用户沟通时,必须始终守住这条底线,原因是:
- 会误伤:非母语写作者、刚学写作的人、翻译腔文本、某些本就高度套路化的类型文(言情、玄幻网文)天然会触发很多「AI 信号」,但完全是人写的。
- 会漏判:AI 被特意提示「写得像人」、或人工把 AI 初稿改透了,信号会被抹掉。
- 人能写得像 AI,AI 能写得像人:文风信号是统计倾向,不是指纹。
- 后果严重:用它去指控学生作弊、判定参赛作品、否定一个作者,可能造成真实伤害。
因此:
- 报告永远给出**「AI 创作度区间 + 置信度」**,而不是一个精确的、像是铁证的数字。
- 措辞用「呈现出与 AI 生成一致的特征」「疑似」「倾向于」,不要用「这就是 AI 写的」。
- 凡是高风险用途(学术诚信、比赛、合同纠纷),明确告知用户:文风分析不能单独作为定论依据,需要配合来源证据(写作过程稿、版本历史、修改痕迹、当场复现能力)。
- 如果文本太短(少于约 300 字)或信号互相矛盾,如实说置信度低,不要硬给高分。
工作流程
仓库内规范入口是 quality/audit/scripts/ 与 quality/audit/references/。本 skill 目录只保留流程说明，审核脚本和参考资料只维护在 quality/audit/。通过 novel-studio skills export --to <dir> 导出时，CLI 会把这些审核资源装配进导出产物的 review/ 目录。
1. 拿到文本
- 若用户上传了文件,先读它(
.txt/.md直接读;.docx/.pdf用对应技能提取纯文本)。 - 若文本已在对话里,直接用。
- 若文本很长(超过约 2 万字),不必逐字精读全文:全文跑量化脚本,同时抽取代表性样本精读——取开头、中段、结尾各 1–2 段,外加 1–2 段对话密集处和 1–2 段描写密集处。在报告里说明抽样方式。
2. 跑量化信号脚本
对全文运行(脚本只依赖标准库,无需联网):
python3 quality/audit/scripts/aigc_value.py <文本文件路径> --target 4
python3 quality/audit/scripts/text_signals.py <文本文件路径>
# 需要结构化结果时加 --json
# 只有用户主动给出腾讯朱雀/AIGC 外部平台抽查值时才原样带入；不得代替用户调用平台
python3 quality/audit/scripts/text_signals.py <文本文件路径> --external-aigc 0.8252
aigc_value.py 会给出本地自研 自研AIGC值(0-1 小数 + 百分比),也就是本文的最终 AI占比；引擎名为 codex-local-aigc-v4。它融合概率曲率、弱语言模型一致性、局部熵/TTR、风格计量、全局/局部语义、叙事动力、内容完整性和分片代理。text_signals.py 会给出句长突发度(CV)、用字多样度、套路措辞、重复长片段、标点习惯、本地与综合风险分。这些是客观可复算的风险证据,但不是作者身份判决。
2.0 内容优先硬原则
审核不是为了让低分文本看起来像合格正文。降 AI 味回改必须先保证剧情、人物、场景和读者读感，再处理检测通过性。以下内容一律视为内容完整性问题，不能作为降 AI 味手段:
- 无语义 OCR 脏码、随机汉字串、字符汤、无语法的长串名词。
- 连续堆叠的稀有神怪名词/怪字串,例如把“魑魅魍魉、饕餮、赑屃、狴犴、貔貅”等连成一行但不承担剧情信息。
- 密集拟声/重复符号长串，且不承担真实场景信息。
- 为制造熵波动而插入的无信息清单、乱码、词表或格式块。
- 数词与正文事实不一致的明显硬伤，例如“几个字：……”后面的实际汉字数量对不上。这类问题必须先改正文，不能被低 AIGC 分数覆盖。
- 顺序词悬空导致句意不明，例如“挂钟先停了”但后文没有“再/然后/随后”的参照。必须改成明确状态或补足顺序关系。
- 状态说明堆成逗号长句，例如“屏幕还亮着，表停在最后一行，批注还在”。这类句子信息没错但读感发硬，必须拆成清楚的动作/视线节奏。
- 开篇、章末或关键转折处硬贴库存明喻，例如“像一根刺”“像一把刀”“像被谁掐住喉咙”。这类句子不一定是 AIGC 证据，但属于读感问题，必须优先改成动作、声音、物件状态或后果。
What ships with it
2 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.
- 10d ago First seen · 296 lines · 242 tokens per session scan A 6525f3c5e237
ai-novel-audit is a skill published in the GitHub repository Xiaoyangy/novel-studio (105 stars, last pushed yesterday), licensed Apache-2.0. It adds 242 tokens to every session and 8,578 once invoked, about $0.0012 per session on Opus 5. A static security scan graded it A with 0 findings. It is 88% identical to smart-search, differing in 347 lines, and is treated as a copy.
Other skills, from other repositories
golang-patterns
Idiomatic Go patterns, best practices, and conventions for building robust, efficient, and maintainable Go applications.
story-import
A tool for turning an existing novel into a structured writing project. It analyzes the book and organizes its characters, settings, plot plans, chapters, and tracking data for later writing.
story-deslop
A writing editor for Chinese web novels that detects writing patterns often associated with AI-generated text and makes the prose feel more natural.
story-long-analyze
A long-form fiction analysis workflow for breaking down a novel’s opening chapters, characters, pacing, turning points, relationships, and overall structure.
story-short-analyze
A Chinese-language guide for analysing short popular web fiction, such as stories published on 番茄, 故事会, or 知乎盐选. It examines the core premise, plot structure, emotional arc, twists, writing techniques, and reader appeal.
story
A Chinese-language routing tool for fiction writers that directs requests to tools for outlining, writing, analysis, reviews, covers, importing novels, browser control, and project management.