story-deslop

story-deslop is a skill for Claude Code, Codex from Xiaoyangy/novel-studio. It costs 62 tokens per session (7,960 once invoked), scanned A, a copy of smart-search, Apache-2.0.

Chinese-language editing guidance for making web fiction sound less like it was produced by a template or AI. It preserves the plot, characters, and important story details while changing the wording and rhythm.

In plain words
What is it for?
Use it to revise Chinese web novels, reduce formulaic phrasing, keep meaningful details, and flag passages that need human review.
Why use it?
Overly polished or repetitive prose can feel unnatural. This helps bring the writing closer to ordinary human storytelling without removing its narrative purpose.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: reads .claude/ paths; mentions Codex; mentions OpenCode.

Good fit Use it to revise Chinese web novels, reduce formulaic phrasing, keep meaningful details, and flag passages that need human review.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/xiaoyangy/novel-studio/story-deslop
Install

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.

Any agent
npx skills add Xiaoyangy/novel-studio --skill story-deslop
Clone the repo
git clone --depth 1 https://github.com/Xiaoyangy/novel-studio

Made for: Claude Code, Codex.

Wrote 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.

agentmods badge for story-deslop

README.md
[![agentmods](https://agentmods.dev/badge/skills/xiaoyangy/novel-studio/story-deslop/github.svg)](https://agentmods.dev/skills/xiaoyangy/novel-studio/story-deslop)
Your own site
<a href="https://agentmods.dev/skills/xiaoyangy/novel-studio/story-deslop"><img src="https://agentmods.dev/badge/skills/xiaoyangy/novel-studio/story-deslop/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.

agentmods 80×15 button for story-deslop

Your own site · 80×15
<a href="https://agentmods.dev/skills/xiaoyangy/novel-studio/story-deslop"><img src="https://agentmods.dev/badge/skills/xiaoyangy/novel-studio/story-deslop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 62 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 7,960 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin 80% copy Near-identical to another mod in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00062 $0.07960
Opus 5 $0.00031 $0.03980
Sonnet 5 $0.00012 $0.01592
Haiku 4.5 $0.00006 $0.00796

Measured 12d ago against content hash e5fe6e4a6e0a, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

story-deslop 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.

The scan reads SKILL.md. This mod also ships 3 executable files (scripts/check-ai-patterns.js, scripts/check-degeneration.js, scripts/normalize-punctuation.js), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

Origin

This is a copy

80% identical to smart-search — 473 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.

skills/story-deslop/SKILL.md · 424 lines

How it starts

The opening of the file, as written. The whole thing — 424 lines — stays where its author put it; the contents beside it link to each section on GitHub.

story-deslop:网文去AI味

你是网文润色专家。你的任务是把 AI 味浓重的网文文本改写自然,降低模板化、书面腔和过度工整感。

核心信念:AI 味的主要问题并非语法错误;更常见的是过度圆滑、工整、解释充分。改写目标是保留剧情功能,同时增加口语、停顿、跳跃和具体动作。


Agent 兼容性:检查专业 agent 是否可用时,按 .claude/agents/{agent}.md.opencode/agents/{agent}.md.codex/agents/{agent}.toml 的顺序查找。Codex 原生子代理调用优先使用同名 agent_type;如果当前 Codex 运行时返回 unknown agent_type 或未暴露 custom-agent registry,必须降级为 solo/direct 执行并报告 fallback。Claude/OpenCode 兼容面保留 subagent_type

核心哲学

原则 1:改味优先,别当改错

AI味不按语法错误处理,也不需要"修正"。它属于风格问题:过于书面化、过于对仗工整、过于面面俱到。去AI味的本质,是把文字从过度工整拉回具体、自然、可读。

原则 2:改最少,效果最大

去AI味不等于重写。目标是改最少的字,让整段文字的"味"变过来。能改一个词就不改一句,能删一句就不重写一段。没有问题的句子尽量保留原句;人名、地名、数字、章节名、专有名词优先保留。

过度去AI味保护

  • 不得整段删除正文内容。如果某段被标记为多处AI味,应逐句修改而非删除整段
  • 删除前必须确认:被删除的内容是否包含伏笔、钩子、角色特征、情节推进等关键信息
  • 如果删除会破坏情节连贯性,改为"降AI重写"而非删除
  • 删除比例上限按 AI 味等级分级:轻度 ≤15%,中度 ≤25%,重度 ≤35%。重度文本可通过“合并重复描写+重写降AI”产生更大字符差,但仍不得整段删除或删掉剧情功能。超过对应比例应在报告中标记超限风险,并输出分段处理方案
  • 如果逐句修改后某段仍不满意,在去AI味报告中标注 [需复核] 而非删除,不计入当前等级的删除比例上限
  • 对于"疑似AI味但不确定"的内容,在去AI味报告中标注 [需复核],而非插入正文

原则 3:保留创作意图

去AI味只改"怎么说",不改"说什么"。剧情、人设、情节走向一概不动;不新增原文没有的情节、设定、关系或时间线。如果原文有逻辑问题,那不是去AI味的活。

原则 4:保留有功能的语气,不保留长停顿符号

去AI味不是把文字全部磨成句号。质问里的 、爆发峰值的少量 可以保留;犹豫、未尽、打断或拖长用动作、短句、换行、逗号或句号重排。正文产物不保留 …… / ——,也要清理无功能的 !!! 和随机标点堆砌。


自然文本基准

去AI味需要知道自然网文文本的特征。以下是从热门网文中提炼的非模板化写作特征,作为对比基准:

自然文本特征(与AI味对比)

维度 自然文本 AI味文本
段落长度 1-3句为主,偶尔1句独占1行 每段4-6句,整齐均匀
对话标签 60%+无标签,用动作替代"说" 几乎每句都有"说道/问道"
情绪表达 动作展示("手在抖") 直接告诉("很紧张")
比喻 生活化("像哈士奇护食") 文学化("如寒冰般")
语气词 "嘤""嘶""靠""行吧" 几乎没有
省略 大量省略,读者自己脑补 面面俱到,生怕读者不懂
排比 偶尔1-2个,从不连续3+ 连续3-5个排比是标配
结尾 动作/对话收尾 总结/升华/感慨收尾

自然表达替换参考

来自大量网文写作研究:

  • 替代"深吸一口气"→ "胸口起伏了一下" / 直接删掉
  • 替代"眼中闪过一丝..."→ "他垂下眼" / "眯起眼"
  • 替代"嘴角勾起一抹..."→ "笑了一下,没到眼底" / "乐了"
  • 替代"仿佛..."→ "像..." / 直接白描
  • 替代"不禁..."→ 直接写动作
  • 替代"缓缓开口"→ "说" / 用动作引出对话

检测流程

Phase 1:AI味扫描

对用户提交的文本做快速扫描,标记AI味浓重的位置:

## AI味检测报告

### 整体评估
- AI味等级:{轻度/中度/重度}
- 主要问题:{1-3 个关键词}

### 问题标记
| 位置 | 类型 | Gate | 原文 | 问题 |
|------|------|------|------|------|
| 第X段 | 禁用词 | A | "眼中闪过一丝..." | 典型AI高频词 |
| 第Y段 | 句式 | B | "...,带着..." | AI惯用句式 |
| 第Z段 | 句式 | B | 连续3句排比 | 过于工整 |
| ... | 心理描写 | C | "他感到..." | 告诉而非展示 |
| 第M段 | 节奏 | D | 段段4-6句、长度均匀 | 整段同节奏 |
| 第N段 | 重复描写 | C/D | 同一动作连续拆写 | 相邻段重复同一瞬间 |
| 第P段 | 解释腔/上帝感 | G | "她不知道的是…" / "演得真好" / "之所以…是因为" | 叙述者跳出角色当下解释/剧透/定性/升华(模式 8) |

> 类型 → Gate 速查:禁用词 = A,句式套路 = B,心理告知 = C,节奏均匀 = D,对话腔调 = E,结尾升华 = F,解释腔/上帝感/安排感 = G,重复描写 = C/D。Phase 2 判定"7 Gate 中 4+ 个有问题"时按 Gate 列计数。

Read the full file on GitHub · 424 lines

Files

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.

Changes

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.

  1. 12d ago First seen · 424 lines · 62 tokens per session scan A e5fe6e4a6e0a

Subscribe to this mod's changes

story-deslop is a skill published in the GitHub repository Xiaoyangy/novel-studio (109 stars, last pushed 3d ago), licensed Apache-2.0. It adds 62 tokens to every session and 7,960 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. It is 80% identical to smart-search, differing in 473 lines, and is treated as a copy.