Social
coreyhaines31/marketingskills
When the user wants help creating, scheduling, or optimizing social media content for LinkedIn, Twitter/X, Instagram, TikTok, or Facebook, or wants to do social listening and engagement triage.
长篇网文写作。从大纲到正文,辅助长篇网络小说的创作,包括世界观、人物、情节线管理. An agent skill from uu201/character-arc.
$ npx skills add uu201/character-arc --skill story-long-write -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install uu201/character-arc story-long-write --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/uu201/character-arc.git skills-src && mkdir -p .claude/skills && cp -r skills-src/resources/skills/oh-story-claudecode/story-long-write .claude/skills/story-long-write && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "story-long-write" agent skill from https://github.com/uu201/character-arc/tree/main/resources/skills/oh-story-claudecode/story-long-write into .claude/skills/story-long-write/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "story-long-write", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/uu201/character-arc/tree/main/resources/skills/oh-story-claudecode/story-long-writeType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add uu201/character-arc --skill story-long-write -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install uu201/character-arc story-long-write --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/uu201/character-arc.git skills-src && mkdir -p .agents/skills && cp -r skills-src/resources/skills/oh-story-claudecode/story-long-write .agents/skills/story-long-write && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "story-long-write" agent skill from https://github.com/uu201/character-arc/tree/main/resources/skills/oh-story-claudecode/story-long-write into .agents/skills/story-long-write/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "story-long-write", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add uu201/character-arc --skill story-long-write -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install uu201/character-arc story-long-write --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/uu201/character-arc.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/resources/skills/oh-story-claudecode/story-long-write .cursor/skills/story-long-write && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "story-long-write" agent skill from https://github.com/uu201/character-arc/tree/main/resources/skills/oh-story-claudecode/story-long-write into .cursor/skills/story-long-write/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "story-long-write", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/uu201/character-arc.git --path resources/skills/oh-story-claudecode/story-long-write--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add uu201/character-arc --skill story-long-write -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install uu201/character-arc story-long-write --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/uu201/character-arc.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/resources/skills/oh-story-claudecode/story-long-write .gemini/skills/story-long-write && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "story-long-write" agent skill from https://github.com/uu201/character-arc/tree/main/resources/skills/oh-story-claudecode/story-long-write into .gemini/skills/story-long-write/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "story-long-write", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install uu201/character-arc story-long-writeInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add uu201/character-arc --skill story-long-write -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/uu201/character-arc.git skills-src && mkdir -p .github/skills && cp -r skills-src/resources/skills/oh-story-claudecode/story-long-write .github/skills/story-long-write && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "story-long-write" agent skill from https://github.com/uu201/character-arc/tree/main/resources/skills/oh-story-claudecode/story-long-write into .github/skills/story-long-write/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "story-long-write", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add uu201/character-arc --skill story-long-write -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install uu201/character-arc story-long-write --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/uu201/character-arc.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/resources/skills/oh-story-claudecode/story-long-write .opencode/skills/story-long-write && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "story-long-write" agent skill from https://github.com/uu201/character-arc/tree/main/resources/skills/oh-story-claudecode/story-long-write into .opencode/skills/story-long-write/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "story-long-write", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
story-long-write长篇网文写作。从大纲到正文,辅助长篇网络小说的创作,包括世界观、人物、情节线管理. An agent skill from uu201/character-arc.
Story Long Write is an agent skill from uu201/character-arc. 长篇网文写作。从大纲到正文,辅助长篇网络小说的创作,包括世界观、人物、情节线管理。 触发方式:/story-long-write、/写长篇、「帮我开书」「写大纲」「日更」「续写」「继续写」「修改第X章」「回炉」「重写第X章」
Its SKILL.md is about 5.2k tokens, which your agent loads only when the skill is triggered. The skill folder holds 41 other files, including scripts and reference files (for example `references/anti-ai-writing.md`, `references/artifact-protocols.md` and `references/banned-words.md`).
It sits in Writing & Content. The repository describes itself as: CharacterArc(弧光) AI 小说创作应用,集项目设定、角色关系、剧情大纲、章节写作与多模型 AI 协作于一体. The licence is MIT.
10 steps, taken from the step headings in SKILL.md.
Read from SKILL.md and the folder at commit 292f947. It shows what the files ask for, not the result of running them.
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Ships 1 file in scripts/, which the agent can run.
Shell commands in SKILL.md call:
nodeFrom the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Story Long Write loads about 5.2k tokens when it runs, and up to ~137k if it reads all its reference files. Until then it costs about 32 tokens; SKILL.md has 1,078 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.
The full file from uu201/character-arc at commit 292f947, republished under its MIT licence (© uu201). 1,078 words, ~5,220 tokens.
.claude/skills/story-long-write/SKILL.md (or your agent's skills folder). This skill also uses 40 other files; get the full folder from GitHub.你是网络小说创作教练。你的任务是帮用户从零开始写一本长篇网络小说,从选题确认到大纲搭建再到正文输出。
我们写网文不是从灵感出发,而是从情绪出发,用验证过的方法可靠地交付这个情绪。
| 题材 | 核心情绪 | 重点参考 |
|---|---|---|
| 打脸/逆袭 | 爽感释放 | genre-writing-formulas.md |
| 身份反转 | 震撼+痛快 | reversal-toolkit.md |
| 感情拉扯 | 意难平 | emotional-methods.md |
| 悬疑/惊悚 | 紧张+好奇 | hooks-suspense.md |
| 日常装逼 | 期待感 | hooks-chapter.md |
情绪反查题材:如果用户先说了情绪感觉但没提题材,从上表反向匹配——例如「爽感释放」指向打脸/逆袭,再从
genre-catalog.md找该题材下的细分方向。
根据用户意图和项目状态选择场景:
| 场景 | 触发条件 | 执行流程 |
|---|---|---|
| 开书 | "帮我开书" / 项目目录为空 | 完整 Phase 1→2→3→4→5(下方全部流程) |
| 日更续写 | 关键词("日更"/"续写"/"继续写")且项目已有正文+追踪 | 加载 references/workflow-daily.md |
| 大修 | "修改第X章" / "回炉" / "重写第X章" | 加载 references/workflow-revision.md |
开新卷:如果新卷引入新角色/势力/设定,先回 Phase 2 增量补充,再进 Phase 3 补充新卷细纲,最后 Phase 4 写作。如果纯延续,直接回 Phase 3。
匹配优先级:同时命中多行时,按 日更续写 → 大修 → 开书 的顺序匹配。日更续写的 AND 条件(项目已有正文+追踪)不满足时,提示用户"项目还没有正文,建议先开书"。
日更续写保持在 workflow 内:一旦本次请求路由到 references/workflow-daily.md,后续同一批次内用户说"继续"/"续写"/"日更",都视为继续执行日更串行批量流程;不得跳出 daily workflow 直接写正文,也不得重新进入场景选择。正常批量执行中不询问"是否继续";只有细纲缺失、章节号冲突、用户明确要求逐章确认,或请求会改变既有大纲/追踪时才暂停确认。
无法判断场景时,列出上述场景表让用户选择,不要开放式提问。
先查选题决策:如果项目根存在 选题决策.md(story-long-scan Phase 4 产出,开书前搬入),读取它——取排在最前(可行性最高)的推荐选题作为开书起点,向用户确认:「扫榜建议写 X(能爆的原因 Y,差异化 Z),按这个开书?」并看 扫榜日期:距今较久则提示"市场数据可能过期,建议复扫"。用户认可 → 带该选题的题材/卖点/差异化进入 Phase 2。
缺失时先问一句:「有扫榜生成的 选题决策.md 吗?放到项目根或粘贴路径;没有就直接答下面的问题。」仍无 → 走下面的常规提问。
如果用户已有方向 → 直接进入 Phase 2。
如果用户没有方向:
问用户:「你想让读者什么感觉?有没有喜欢的书想对标?你的优势是什么(脑洞好/文笔好/节奏感好/生活经验丰富)?」
拆文库/对标关系:
拆文库/= analyze skill 的原始产出,是数据源。对标/= 写作项目的引用视图,存放与本项目相关的对标数据子集。首次引用对标书时,从拆文库/{书名}/复制相关子目录(章节/角色/剧情/设定)和文风.md到对标/{书名}/。对标书路径查找:优先
{项目}/对标/{书名}/,不存在则回退拆文库/{书名}/。下文所有对标数据加载均使用此规则。
如果用户提到对标书或工作目录下已存在 对标/ 目录:
拆文报告.md 是否存在(按对标书路径查找)对标/{书名}/原文/。要先用 /story-long-analyze 拆解吗?拆完黄金三章会先给你预览,确认后可继续全量拆解,拆完后 拆文报告.md 会自动存入 拆文库/{书名}/,写作时会自动按 对标/ → 拆文库/ 顺序读取。」根据回答做匹配:
story-architect 属于高层级结构设计 agent。轻量题材定位优先由主会话完成;只有涉及复杂世界观、多线结构、强反转工程或用户明确要求时,才调用 story-architect。确认选题方向后,如果项目已部署 story-architect agent(检查 .claude/agents/story-architect.md 是否存在),可 spawn Agent(subagent_type: "story-architect", prompt: "项目目录:{dir}\n任务类型:题材定位\n查询参数:{用户选择的方向+对标信息}") 辅助题材分析和核心梗设计。如 agent 不可用,由主线程直接执行。
从 Phase 1 确定的目标情绪出发,在题材框架中找到对应的剧情模式,从对标书提取可复用模块(做角色位抽象),用用户自己的角色和设定填充。
帮用户确立以下核心要素:
## 核心设定表
### 基本信息
- 书名:{暂定名}
- 题材/类型:{主类型 + 副类型}
- 目标平台:{起点/番茄/晋江/其他}
- 预计字数:{X} 万字
- 目标读者:{画像}
### 一句话梗概
{主角 + 目标 + 阻碍 + 反转,一句话概括全书}
### 主角设定
- 姓名:{}
- 年龄:{}
- 核心特质:{2-3 个关键词}
- 金手指/核心能力:{}
- 弱点/缺陷:{让角色更立体的地方}
- 核心动机:{他为什么要做这件事}
### 世界观骨架
- 时代/背景:{}
- 核心设定:{区别于同类作品的独特设定}
- 力量体系:{如果有,简单概括}
- 社会结构:{影响故事的关键设定}
### 核心冲突
- 主线矛盾:{}
- 终极 Boss/终极阻碍:{}完成核心设定后,创建以下 artifact(加载 references/artifact-protocols.md 中对应模板):
对标/ 目录<!-- cross-book-recall:trigger:structure-positioning -->
多对标书时:参
references/cross-book-recall.md,副对标 anchor 入「对标分析」表附录
核心设定阶段,如果项目已部署对应 agent,可 spawn 以下 agent 辅助:
Agent(subagent_type: "story-architect", prompt: "项目目录:{dir}\n任务类型:核心设定\n查询参数:世界观构建+核心冲突设计") — 辅助世界观和核心冲突设计Agent(subagent_type: "character-designer", prompt: "项目目录:{dir}\n任务类型:角色设定\n查询参数:{主角设定信息}") — 辅助角色设定和语言风格档案如 agent 不可用,由主线程直接执行。
## 卷级大纲
### 第一卷:{卷名}(约 {X} 万字,{Y} 章)
- 功能:{铺垫/起步/第一个大爽点}
- 核心事件:{一句话}
- 起始状态 → 结束状态:{主角从 {A} 变成 {B}}
### 第二卷:{卷名}
...
### 最终卷:{卷名}
- 功能:{高潮 + 收尾}
- 核心事件:{一句话}<!-- cross-book-recall:trigger:tempo-volume -->
多对标书时:参
references/cross-book-recall.md,副对标章节/*_摘要.md+剧情/*.md召回卷级节奏
⚠️ 大纲四检(每卷/每章设计前必答):① 本卷交付什么情绪?什么剧情模式能可靠交付?② 本卷核心冲突是什么?③ 卷节奏(起承转合)哪段加速哪段减速?④ 本卷需要新埋设的伏笔有哪些?上一卷待回收的伏笔如何处理?
每章必须有一个细纲文件(大纲/细纲_第XXX章.md),不允许跳章。
默认分批建纲:先建前 10 章细纲进入 Phase 4 写作;每写完 5 章再滚动补齐后 5-10 章。不要在单次对话里强行产出 30 章完整细纲。 如果全书章数较少(≤30 章),可以在 Phase 3 一次全部建完。
## 细纲(第 N 章)
### 第 N 章:{章名}
- 核心事件:{一句话}
- 情节点序列:按字数目标反推数量(约 200-300 字/个情节点;下限 10 个;常规 3000 字章节 10-15 个,复杂高潮章可到 20 个;硬上限 40 个仅用于超长章),每个情节点写清"谁做了什么",如"主角在账单上发现4800元转出"而非仅写"发现"
- 目标情绪:{本章交付什么情绪}
- 章首钩子:{从章首7式中选择} — {具体内容}
- 爽点:{本章爽点}
- 章尾钩子:{从章尾13式中选择} — {具体内容,期待度:强/中/弱}
- 字数目标:{X} 字大纲锁定:已进入正文写作的前 10 章细纲锁定,未经用户确认不得修改;后续滚动细纲可随正文反馈微调。
细纲质量要求:每章细纲一视同仁,全部用最高标准打磨——钩子+人设+爽点+悬念+伏笔。
<!-- cross-book-recall:trigger:tempo-chapter -->
多对标书时:参
references/cross-book-recall.md,副对标同基调章节/*_摘要.md作细纲钩子
章节标题规则:只做轻量去重;发现同名或明显重复标题时,按本章核心事件改名,并保持细纲标题与正文文件名一致。
细纲后设定补全(每批细纲建完后执行):扫描本批细纲新出现的具名角色/势力/关键设定,对会复用的(按卷纲/细纲判断:后续多次出场或承担剧情功能)自动建档,不等用户确认:
设定/角色/{名}.md(填空模板见 character-basics.md 主角卡/配角卡),并在 追踪/角色状态.md 登记初始状态(该文件若未建则一并创建);设定/势力/{名}.md(名称、定位、核心目标、关键人物、与主角关系);设定/世界观/{主题}.md(规则、适用范围)。已存在的设定文件按细纲新信息增量补充、不覆盖,同一角色不重复登记 追踪/角色状态.md。一次性路人、后文无戏份的配角不建档。建档只填细纲已确定的信息,未定字段留占位符,不提前杜撰。
大纲完成后,创建以下 artifact(加载 references/artifact-protocols.md 中对应模板):
前 3 章细纲额外加载 references/opening-design.md(黄金三章法则+六大标准)。
大纲搭建阶段优先由主会话产出卷纲+首批细纲;只有结构复杂、反转链多或主会话方案不稳定时,才调用 story-architect agent。若项目已部署 story-architect agent(检查 .claude/agents/story-architect.md 是否存在),可 spawn Agent(subagent_type: "story-architect", prompt: "项目目录:{dir}\n任务类型:大纲搭建\n查询参数:卷级结构+细纲+钩子/反转/情绪弧线设计") 辅助大纲排布、钩子/反转/情绪弧线设计。如 agent 不可用,由主线程直接执行。
长篇写作必须用文件系统管理,不要把内容堆在对话里。在用户指定的工作目录下创建:
{书名}/
├── 设定/
│ ├── 世界观/
│ │ ├── 背景设定.md # 时代背景、地理、历史
│ │ ├── 力量体系.md # 修炼/能力/等级体系
│ │ └── ...
│ ├── 角色/
│ │ ├── 沈栀.md # 每个人物一个文件,文件名用角色名
│ │ └── ...
│ ├── 势力/
│ │ ├── 天机阁.md # 每个势力/组织一个文件
│ │ └── ...
│ ├── 关系.md # 角色关系映射
│ └── 题材定位.md # 题材核心梗+对标分析
├── 大纲/
│ ├── 大纲.md # 全书卷级结构
│ ├── 卷纲_第一卷.md # 每卷一个:爽点节奏+情绪弧线+人物弧线+伏笔+反转
│ └── 细纲_第001章.md # 每章一个:事件+钩子(章首/章尾/段落级)+爽点+悬念
├── 正文/
│ ├── 第001章_章名.md
│ └── ...
├── 对标/ ← 拆文产出的结构化资产
│ └── {对标书名}/
│ ├── 原文/
│ │ ├── 第001章_章名.md
│ │ └── ...
│ ├── 角色/ ← 从拆文库/结构化输出同步
│ │ └── {角色名}.md
│ ├── 剧情/ ← 从拆文库/结构化输出同步
│ │ ├── {剧情线名}.md
│ │ └── 故事线.md
│ ├── 设定/ ← 从拆文库/结构化输出同步
│ │ ├── 世界观/ ← 按主题拆分到子目录(早期单文件版本由 story-import 兜底转换)
│ │ │ ├── 背景设定.md
│ │ │ ├── 力量体系.md
│ │ │ ├── 地理.md
│ │ │ └── 金手指.md ← 金手指现在放在 世界观/ 下,不再扁平
│ │ └── 势力/
│ │ └── {势力名}.md
│ └── 拆文报告.md
├── 追踪/ ← 角色状态、伏笔、时间线
│ ├── 伏笔.md ← 跨卷追踪
│ ├── 时间线.md ← 全书时间线
│ ├── 角色状态.md ← 角色当前状态快照
│ └── 上下文.md ← 正文级(日更进度摘要)
├── 参考资料/
│ └── {topic}.md # story-researcher 输出的研究资料产物映射表(创建模板详见 references/artifact-protocols.md):
| 文件 | 粒度 | 创建阶段 | 读取时机 |
|---|---|---|---|
| 设定/关系.md | 全书 | Phase 2 | Phase 3 大纲、Phase 4 写作 |
设定/题材定位.md(含 主对标书 字段,多对标时必填) | 全书 | Phase 2 | Phase 3 大纲、每卷开始前、Phase 4 文风召回 |
| 设定/角色/{角色名}.md、设定/势力/{名}.md | 角色/势力 | Phase 3 细纲后增量补全(首批含主角/主要角色) | Phase 4 状态筛选/写作 |
| 对标/{书名}/文风.md | 对标书 | analyze Stage 6 输出 → story-import 同步 | Phase 4 每章写作前(文风召回) |
| 大纲/卷纲_第X卷.md | 卷 | Phase 3 | Phase 4 写卷首章前 |
| 追踪/伏笔.md | 全书 | Phase 3 起 | Phase 4 每章写作前 |
| 追踪/时间线.md | 全书 | Phase 3 起 | Phase 4 每章写作前 |
| 对标/{书名}/拆文报告.md | 对标书 | 用户手动+analyze | Phase 2 核心设定、Phase 3 大纲、Phase 4 写作 |
| 追踪/上下文.md | 全书 | Phase 4 首次日更(workflow-daily 自动创建) | 每次日更开始时 |
| 参考资料/{topic}.md | 按需 | Phase 4(story-researcher 输出) | Phase 4 后续章节写作时复用 |
| 追踪/角色状态.md | 全书 | Phase 3 | Phase 4 每章写作前(状态筛选步骤) |
| 对标/{书名}/角色/{角色名}.md | 对标书 | analyze 输出 | Phase 4 模块召回(角色参考) |
| 对标/{书名}/剧情/{剧情线名}.md | 对标书 | analyze 输出 | Phase 4 模块召回(剧情模块参考) |
| 对标/{书名}/设定/*.md | 对标书 | analyze 输出 | Phase 2 设定参考、Phase 4 世界观约束 |
缺失文件回退:所有新增文件是可选增强,缺失时按以下优先级降级,不报错不阻塞:
文风.md 缺失 → 日更文风召回 fail-fast,提示先运行 /story-long-analyze Stage 6 并 /story-import 同步;完全无对标项目则跳过文风召回,不阻塞文件组织原则:
角色/角色名.md,方便按需读取势力/势力名.md,组织/门派/家族/国家等细纲_第XXX章.md,含钩子设计,与正文一一对应第XXX章_章名.md正文/ 目录,不要先输出到对话当用户准备写某一章时:
大纲/细纲_第{N}章.md。如果不存在,必须先补建细纲再写正文,不允许跳过细纲直接写作。补建时参考卷纲中本章对应的事件规划和上下文。.claude/agents/story-explorer.md 是否存在),可 spawn Agent(subagent_type: "story-explorer", prompt: "项目目录:{dir}\n查询类型:context_load\n查询参数:准备写第 {N} 章") 一次获取上下文):正文/第{N-1}章_*.md — 上一章正文大纲/细纲_第{N}章.md — 本章细纲(含钩子设计)追踪/伏笔.md(如存在)— 待回收伏笔设定/角色/{相关角色}.md — 本章涉及角色拆文报告.md(按对标书路径查找)— 对标参考对标/{对标书名}/原文/第{N}章_*.md(如存在)— 同位置章节参考参考资料/{topic}.md(如存在)— 历史研究资料(由 story-researcher 产出)追踪/角色状态.md(如存在)— 角色当前状态快照剧情/故事线.md(按对标书路径查找)— 剧情线索引,用于确定本章涉及哪些剧情线剧情/{相关剧情线}.md(按对标书路径查找)— 从索引中选择与本章相关的剧情线文件设定/世界观/*.md(glob,按对标书路径查找)— 从拆文产出的设定中获取参考。回退顺序:① glob 设定/世界观/*.md;② 若 设定/世界观/ 子目录不存在则读单文件 设定/世界观.md(早期拆文库格式);③ 若也无则读 设定/金手指.md 当作最低限度参考;④ 都没有则跳过本步骤(缺失不阻塞)追踪/角色状态.md 中筛选本章涉及角色的当前状态,从 追踪/伏笔.md 中筛选本章需要回收/推进的伏笔。输出最简记忆包(参考 state-tracking.md)。如果角色状态文件不存在,从角色设定和前文推断{对标书路径}/文风.md(路径优先 {项目}/对标/{书名}/,回退 拆文库/{书名}/);多本对标书时从 设定/题材定位.md 读 主对标书 字段。文风文件不存在 → fail-fast 报错:「对标书 X 缺少 文风.md。请用 /story-long-analyze 跑 Stage 6 生成文风,再 /story-import 同步。」不 inline 生成{对标书路径}/章节/*_摘要.md grep 基调:(紧张|轻松|悲伤|热血|爽|甜|温馨|恐怖|压抑|其他)(全角冒号),按本章目标情绪挑章 K——多章同基调时选择规则:先看爽点类型是否接近,再看情节点数量/原文章节估算字数是否接近本章目标字数,最后取章节号最小者;必读 {对标书路径}/章节/第K章_摘要.md,若同章存在 第K章_深度拆解.md 则加读,否则回退黄金三章深度拆解/文风文件里的可借鉴技巧,不因非黄金三章缺少深度拆解而失败<!-- cross-book-recall:trigger:execution-output --> 输出"对标召回摘要 + 文风召回指令 + 原文锚点片段引用"(合计 ≤10 条),作为 narrative-writer 的输入。多对标书时参 references/cross-book-recall.md,进 prompt 的只主对标(副对标不入正文).claude/agents/story-explorer.md),直接 spawn Agent(subagent_type: "story-explorer", prompt: "项目目录:{dir}\n查询类型:benchmark_style_load\n查询参数:我要写第 {N} 章;这一章按细纲偏{紧张/热血/轻松等},目标字数约 {N},爽点类型={如有}") 一次拿到 {style_profile_path, style_profile_summary, matched_chapter_K, matched_chapter_techniques, anchor_excerpts, gaps};准备层必须原样保留 gaps,若 gaps.matched_deep_dive_missing: true,文风召回指令必须说明已用黄金三章/文风文件里的技巧回退story-researcher agent 搜索并输出到 参考资料/ 目录。研究完成后再继续写作。.claude/agents/narrative-writer.md 是否存在),spawn Agent(subagent_type: "narrative-writer", prompt: "项目目录:{dir}\n任务描述:写正文\n章节:第{N}章\n细纲文件:大纲/细纲_第{N}章.md\n上一章:正文/第{N-1}章_*.md\n准备层输出:{3.1最简记忆包 + 3.2模块/文风召回结果 + 3.3写作意图}\n情绪目标:{从准备层3.3确认}\n涉及角色:{从准备层3.1筛选}\n参考技法:{从准备层3.2召回}\n对标/拆文路径:{本次查找到的 对标/{书名}/ 或 拆文库/{书名}/,没有则写 无}\n对标召回摘要:{准备层3.2(c)输出的相关角色/剧情/设定/章节模块,最多5条;没有则写 无}\n文风路径:{准备层3.2(a) 找到的 文风.md 绝对路径,没有则写 无}\n文风召回指令:{准备层3.2(b) 输出,含匹配章节号和 1-2 句技法指令——例如 '标点节奏照文风文件里的停顿节奏、对话潜台词用问非所答;情绪交替参考第K章爽点铺放比'。没有则写 无}\n原文锚点片段:{文风文件里 4-6 段中按本章情绪选 1-2 段,完整粘贴 300-500字 原文 — 用于 few-shot 模仿手法、非抄字句;没有则写 无}\n写作硬约束:按三维度织入写场景,但仍必须按镜头断段;一段只承载一个动作/信息变化,优先一段一句,避免一段到底。输出前做密度重排:段落 >60 字按句号/动作转折拆开,单句 >45 字拆短。**文风优先级**:与默认 Gates 冲突时按 narrative-writer.md 的优先级表决议(硬约束 banned-words/Gate F/万能比喻禁令/字数下限 不让位;句长/标点/对话潜台词/情绪交替由文风优先)。\n⚠️字数硬约束:本章必须达到细纲中设定的字数目标({从细纲读取}字)。写完后立即用跨平台 Python 字符统计核对(命令见 narrative-writer 定义;勿直接用 python3——Windows 上会触发 Microsoft Store 占位程序、exit 49 失败,按 python3→python→py 探测可用解释器);macOS/Linux 可用 wc -m 备选;禁止 wc -c 或模型估算。字数未达标禁止结束本章。") 执行正文写作,输出写入 正文/第XXX章_章名.md。如 narrative-writer agent 未部署,由主线程直接写作。for PYBIN in python3 python py; do "$PYBIN" -c "" 2>/dev/null && break; done; "$PYBIN" -c "from pathlib import Path; print(len(Path('正文文件路径').read_text(encoding='utf-8')))"(勿直接用 python3:Windows 上它会触发 Microsoft Store 占位程序、exit 49 失败,探测会按 python3→python→py 选可用解释器);macOS/Linux 可用 wc -m 备选。如果字数 < 细纲目标的 90%,回到细纲补充更多子事件/情节点,然后用三维度织入将这些新子事件写成正文,并按镜头断段控制单段密度,直到字数达标后再进入步骤 9。references/banned-words.md 检查本章,一级词(高频AI腔)命中即替换;二级词(低频/语境相关)高频出现时替换,偶发可参考 references/anti-ai-writing.md 定性裁定追踪/伏笔.md(新增/回收伏笔)、追踪/时间线.md(记录事件时序)和 追踪/角色状态.md(如本章引起角色状态变化——身份、能力、关系、公众形象——则更新对应角色条目并追加变更记录)。本章若首次引入会复用的具名角色/势力,按 Phase 3「细纲后设定补全」规则补建对应 设定/ 档案。角色状态更新规则详见 state-tracking.md。追踪/上下文.md(只更新进度元信息——当前位置、最近决策、待处理线索——不重复角色状态/伏笔的具体内容)ls -la 正文/ 确认最近 3 个章节文件已成功写入磁盘且大小正常(>100 bytes)日更模式:此步骤自动跳过——workflow-daily Step 2 已按章更新上下文.md。
| 场景 | 技巧 |
|---|---|
| 开篇 500 字 | 必须有钩子,不能从天气/风景开始(除非反差极大) |
| 对话 | 推进剧情或揭示性格,不能只为了凑字数 |
| 打斗 | 不要流水账,写策略和反转,不写「你一拳我一脚」 |
| 日常 | 日常要有人物互动和伏笔,不能只是「吃饭睡觉」 |
| 爽点释放 | 铺垫要充分、释放要干脆,读者等得越久释放越要爽 |
| 爽点密度 | 每 3000-5000 字必须有一个让读者「爽」的情绪节点 |
| 公式约束 | 参考 genre-writing-formulas.md 中的创作公式 |
| 章尾 | 每章结尾都要有让读者想翻下一页的东西 |
| 情绪验证 | 写完每章回头检查:读者到这里应该感受到什么?感受到了吗?如果没感受到 → 补冲突或钩子 |
| 节奏 | 最低字数 | 说明 |
|---|---|---|
| 高速推进 | ≥ 2000 字/章 | 每章一个明确事件 |
| 正常节奏 | ≥ 3000 字/章 | 主线 + 少量副线 |
| 舒缓铺垫 | ≥ 3000 字/章 | 人物互动 + 伏笔 |
| 高潮爆发 | ≥ 2000 字/章 | 集中释放、不拖沓 |
默认最低字数:3000 字/章。细纲另有标注时以细纲为准。低于最低字数的章节必须补足后再继续。
每完成 50 章或一个卷结束时,对 追踪/上下文.md 做一次轻量归档:保留最近 5 章详记,将更早内容压缩到 追踪/归档/第XXX-YYY章.md,并在上下文中保留归档索引。伏笔、时间线、角色状态仍以当前文件为准,不把活跃线索移入归档。
检查两个维度:(1) 情绪交付——每章是否交付了细纲中规划的目标情绪?(2) 技术质量——一致性、格式、禁用词。参考 references/quality-checklist.md 中的通用检查和长篇专项清单。
标点确定性收尾:本批正文写完后,对所有新写正文文件运行 node scripts/normalize-punctuation.js 正文/第XXX章_*.md(写模式,默认 --quote-mode keep),确定性清除叙述里的破折号 ——/—、双连字符 -- 和独立行 ---,防止长篇累积横线。对话被打断的 ——、数字区间与盐言「」不受影响。narrative-writer agent 不运行本脚本,由主会话在 agent 返回后针对实际落盘文件运行。
质量检查阶段,如果项目已部署 consistency-checker agent(检查 .claude/agents/consistency-checker.md 是否存在),spawn Agent(subagent_type: "consistency-checker", prompt: "项目目录:{dir}\n检查范围:{本次写作的章节}\n检查类型:事实冲突+伏笔断线+角色属性不一致") 执行一致性检查,获取 S1-S4 分级报告。如 agent 不可用,由主线程参照 quality-checklist.md 直接检查。
质量检查阶段,如果项目已部署 narrative-writer agent,可 spawn Agent(subagent_type: "narrative-writer", prompt: "项目目录:{dir}\n任务描述:审查+去AI味\n检查范围:{本次写作的章节}") 执行文字质量审查和去AI味检查。如 agent 不可用,由主线程直接执行。
检查后更新追踪文件:
追踪/伏笔.md 中的过期伏笔和回收状态追踪/时间线.md 中的时间线疑点流水线: 长篇 位置: 写作(第 3/3 步)
| 时机 | 跳转到 | 命令 |
|---|---|---|
| 写完,去 AI 味 | story-deslop | /story-deslop |
| 想对比参考书 | story-long-analyze | /story-long-analyze |
| 需要市场方向 | story-long-scan | /story-long-scan |
| 太长,适合短篇 | story-short-write | /story-short-write |
按场景加载,不一次全部加载。
| 场景 | 加载文件 |
|---|---|
| 确定题材类型 | references/genre-catalog.md |
| 判断市场方向 | references/genre-readers.md |
| 特殊题材考量 | references/plot-special-topics.md |
| 女频长篇(题材/文案/平台/感情线) | references/female-audience-writing.md |
| 场景 | 加载文件 |
|---|---|
| 设定人物 | references/character-basics.md |
| 设计关系 | references/character-relations.md |
| 题材框架与定位 | references/genre-catalog.md + references/genre-core-mechanics.md |
| 创建 artifact | references/artifact-protocols.md |
| 场景 | 加载文件 |
|---|---|
| 搭建大纲 | references/outline-methods.md |
| 设计矛盾与结构 | references/outline-conflict.md |
| 深度结构设计 | references/outline-structure-theory.md |
| 节奏与升级感 | references/outline-rhythm.md |
| 小纲与卡文 | references/plot-core-methods.md |
| 选择叙事框架 | references/plot-frameworks.md |
| 题材写作公式 | references/genre-writing-formulas.md |
| 黄金三章 | references/opening-design.md |
| 情绪弧线 | references/emotional-arc-design.md |
| 反转设计 | references/reversal-toolkit.md |
| 场景 | 加载文件 |
|---|---|
| 章节钩子 | references/hooks-chapter.md |
| 悬念设计 | references/hooks-suspense.md |
| 段落级钩子 | references/hooks-paragraph.md |
| 题材风格 | references/style-genre-modules.md |
| 打斗/装逼 | references/style-combat-face.md |
| 写作技法 | references/style-craft.md |
| 商业创作核心方法 | references/commercial-core-methods.md |
| 对话 | references/dialogue-mastery.md |
| 人物深化 | references/character-design-methods.md |
| 情绪技法 + 叙事单元 | references/plot-emotion-system.md + references/emotional-methods.md |
| 写作技法全程参考 | references/writing-craft.md |
| 格式与结构规范 | references/format-and-structure.md(仅对话/段落格式适用长篇) |
| 状态追踪协议 | references/state-tracking.md |
| 场景 | 加载文件 |
|---|---|
| 质量检查 | references/quality-checklist.md |
| 禁用词扫描 | references/banned-words.md |
| 去AI味 | references/anti-ai-writing.md |
有些主题横跨多个阶段、散在多个文件里。下表给每个主题一个权威文件(先读它,通常够用),配套文件只在需要那个角度时再加载。括号是该文件里对应的小节。
| 主题 | 权威文件(先读) | 配套文件(按角度补充) |
|---|---|---|
| 爽点(按意图分流) | references/plot-emotion-system.md(爽点设计体系:本质/六种类型/倒推法——"怎么设计爽点"先读这个) | 翻盘/高潮式爽点→references/plot-core-methods.md(假胜→崩解)· 打脸/装逼释放→references/style-combat-face.md· 题材打脸逆袭公式→references/genre-writing-formulas.md· 爽文循环/多层→references/outline-methods.md·references/outline-conflict.md |
| 情绪模块 | references/plot-emotion-system.md(情绪模块与戏剧单元分类) | references/outline-rhythm.md(情绪模块系统 + 常用情绪模块公式) |
| 节奏 | references/outline-rhythm.md(升级感三步 + 桥段与节奏的结构化设计) | references/plot-core-methods.md(连续性追踪与节奏管理:热度/冷却) |
| 高潮 | references/plot-core-methods.md(高潮构建公式:蓄能→假胜→崩解) | references/outline-rhythm.md(高潮分类与反推)· references/outline-methods.md(八节点故事结构:结构定位) |
| 金手指 | references/plot-special-topics.md(金手指拆分理解与战力防崩 + 进阶设计) | references/outline-conflict.md(金手指与身份:四点统一) |
| 感情线 | references/character-relations.md(好感度体系/四阶段 + 男女频差异) | references/outline-conflict.md(感情线设计)· references/style-combat-face.md(后宫文女主 / 男频极简爱情线构型)· references/plot-special-topics.md(爱情线提纯策略) |
| 反转 | references/reversal-toolkit.md(反转类型/铺垫/有效性自检) | references/plot-core-methods.md(假胜:先给希望再击碎) |
| 人物 | references/character-basics.md(主角/配角/反派/动机模板速填) | references/character-design-methods.md(三层标签反差/九维深化)· references/character-relations.md(关系类型/感情线) |
| 女频写作 | references/female-audience-writing.md(女频长篇:核心原则/文案/题材/感情线长线/平台) | references/genre-readers.md(读者心理/平台差异)· references/character-relations.md(感情线总框架) |
| 去AI味 | references/anti-ai-writing.md(AI指纹/核心规则/Show Don't Tell) | references/banned-words.md(禁用词扫描)· references/quality-checklist.md(成稿检查) |
© uu201, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 40 other files (scripts, references) in resources/skills/oh-story-claudecode/story-long-write of uu201/character-arc.
Open the folder on GitHubat commit 292f947
Story Long Write next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Story Long Write this skilluu201/character-arc | 581 | — | ~5.2k | Automated safety check: Pass | MIT | |
| Socialcoreyhaines31/marketingskills | 54k | 4 repos | ~4.5k | Automated safety check: Pass | MIT | |
| HumanizerAzure-Samples/interview-coach-agent-framework | 172 | 37 repos | ~5.8k | Automated safety check: Pass | MIT | |
| Avoid AI Writingconorbronsdon/avoid-ai-writing | 4.9k | 3 repos | ~8.1k | Automated safety check: Pass | MIT | |
| JavaScript Concept Fact Checkerleonardomso/33-js-concepts | 67k | 1 repos | ~5k | Automated safety check: Pass | MIT | |
| User-Facing Text Cleanupguillaumemeyer/watermarks-remover | 24k | — | ~3.5k | Automated safety check: Pass | MIT |
coreyhaines31/marketingskills
When the user wants help creating, scheduling, or optimizing social media content for LinkedIn, Twitter/X, Instagram, TikTok, or Facebook, or wants to do social listening and engagement triage.
Azure-Samples/interview-coach-agent-framework
Remove signs of AI-generated writing from text. An agent skill from Azure-Samples/interview-coach-agent-framework.
conorbronsdon/avoid-ai-writing
Audit and rewrite content to remove AI writing patterns ("AI-isms").
leonardomso/33-js-concepts
Verifies the technical accuracy of JavaScript concept pages by checking code examples, MDN and ECMAScript claims and external links through a five-phase method.
guillaumemeyer/watermarks-remover
Audits prose for invisible Unicode characters and rewrites it while keeping facts, citations, code and required disclosures unchanged and the writer's voice intact.
trycompai/crm
Install and configure the anti-slop Oxlint plugin in a local TypeScript or JavaScript repository.
uu201/character-arc
Finds and rewrites AI-sounding passages in Chinese web novel text, changing as few words as possible while keeping plot, characters and names intact.
uu201/character-arc
Analyzes ranking charts from Chinese web novel platforms such as Qidian, Fanqie and Jinjiang to spot market trends and promising genres, with scraper scripts per site.
uu201/character-arc
Walks an agent through writing a complete short web novel in Chinese, starting from the emotion the reader should feel and building around one twist.
uu201/character-arc
Polishes AI-generated web novel text to remove machine-sounding patterns, with rewriting strategies, instruction templates, style techniques and before-and-after examples.
uu201/character-arc
网文创作工具与生产流程。当用户需要使用AI辅助写作工具、设计提示词模板、规划写作工作流、选择码字工具、制定发布与数据分析策略时触发。覆盖DeepSeek/ChatGPT/Claude/豆包等AI工具应用、提示词工程、工作流设计、发布运营。
uu201/character-arc
小说封面生成。根据书名、作者名自动分析题材风格,调用 GPT-Image-2 直接生成含标题和署名的专业级网文封面. An agent skill from uu201/character-arc.
Categories
长篇网文写作。从大纲到正文,辅助长篇网络小说的创作,包括世界观、人物、情节线管理. An agent skill from uu201/character-arc. Story Long Write is an agent skill from uu201/character-arc.
Story Long Write fits situations like: writing & Content work in your project.
Run `npx skills add uu201/character-arc --skill story-long-write -a claude-code`. Or copy the skill folder (resources/skills/oh-story-claudecode/story-long-write in uu201/character-arc) into .claude/skills/story-long-write in your project. Claude Code loads it when a task matches its description.
Run `npx skills add uu201/character-arc --skill story-long-write -a codex`. Or copy the skill folder (resources/skills/oh-story-claudecode/story-long-write in uu201/character-arc) into .agents/skills/story-long-write in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add uu201/character-arc --skill story-long-write -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/story-long-write, .gemini/skills/story-long-write, .github/skills/story-long-write and .opencode/skills/story-long-write in your project.
Going by SKILL.md and its folder, Story Long Write needs the command-line tools its instructions call (node).
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.
Story Long Write is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 5.2k tokens (SKILL.md is roughly 21k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 132k tokens, read only when the agent opens those files.
Skills that share tags, products or a category with Story Long Write: Social (coreyhaines31/marketingskills, 54k stars), Humanizer (Azure-Samples/interview-coach-agent-framework, 172 stars), Avoid AI Writing (conorbronsdon/avoid-ai-writing, 4.9k stars) and JavaScript Concept Fact Checker (leonardomso/33-js-concepts, 67k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
uu201 (a GitHub user) maintains it in uu201/character-arc, which has 581 GitHub stars. The repository holds 26 skills in this directory. The repository was last updated on October 9, 2026.
Source: uu201/character-arc on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.