claude-screenwriting
CLAUDE.md9 starsChanged 6 months agoMITView on GitHub
这是一个 Obsidian vault + Claude Code 编剧工作环境。Claude 在此 vault 中充当 AI 编剧助手,基于 Vogler/McKee/Save the Cat 理论维护结构化故事 wiki。 详见 skills/screenwriting/references/ 目录。 所有新页面使用 _templates/ 目录下的模板创建: - beat-card.md — Beat 节拍卡 - character-card.md — 角色卡(3 Tab 结构) - scene-card.md — 场景卡 - logline.md — 故事基因 - story-index.md — 主索引
In this repository8 other files it gives its agents.
- screenwriting-beatsSkill
- screenwriting-characterSkill
- screenwriting-exportSkill
- screenwriting-ingestSkill
- screenwriting-lintSkill
- screenwriting-refactorSkill
- screenwriting-reimagineSkill
- screenwritingSkill
The file
Files
Claude-Screenwriting 项目指令
这是什么
这是一个 Obsidian vault + Claude Code 编剧工作环境。Claude 在此 vault 中充当 AI 编剧助手,基于 Vogler/McKee/Save the Cat 理论维护结构化故事 wiki。
核心规则
- 所有故事结构操作必须基于编剧理论 — 不能随意发明结构,必须引用框架
- 先读后写 — 任何写入操作前必须先读取当前状态
- Diff 可见 — 结构变更先展示 diff,用户确认后再写入
- 来源可追溯 — 每个 insight 都能通过 backlink 追溯到素材来源
- 不编造 — 不确定的信息标记为 uncertain,不要凭空填充
Vault 结构
wiki/
├── index.md # 故事主索引
├── hot.md # 热缓存(当前上下文摘要)
├── log.md # 操作日志
├── beats/ # 15 Beat 页面
├── characters/ # 角色卡页面
├── sources/ # 素材摘要页
└── meta/ # logline / genre / synopsis可用命令
| 命令 | 功能 |
/screenwriting | 初始化 vault 或查看当前状态 |
/screenwriting-ingest [file] | 从素材文件抽取 beats/characters/meta + 生成 canvas |
/screenwriting-beats | 查看/编辑/生成 15-beat 表 |
/screenwriting-character [name] | 查看/创建/编辑角色卡 |
/screenwriting-refactor [direction] | 重构故事结构(带 diff) |
/screenwriting-reimagine [direction] | 基于现有版本创造性重构,生成全新版本 |
/screenwriting-lint | 运行结构健康度诊断 |
/screenwriting-export [format] | 导出大纲/人物表/剧本骨架 |
理论框架
- Vogler 8 原型: Hero, Mentor, Ally, Shadow, Shapeshifter, Guardian, Herald, Trickster
- McKee: 矛盾(surface/truth)、压力选择、Wants vs Needs、人物维度
- Save the Cat: 15 beats、四大支柱、碎玻璃、六大缺陷
详见 skills/screenwriting/references/ 目录。
写入规范
- Beat 文件命名:
beat-{XX}-{中文名}.md(XX 两位数字) - 角色文件命名:
{角色名}.md - 所有新页面末尾必须有 source backlink
log.md为 append-only,每次操作追加一条记录
模板
所有新页面使用 _templates/ 目录下的模板创建:
beat-card.md— Beat 节拍卡character-card.md— 角色卡(3 Tab 结构)scene-card.md— 场景卡logline.md— 故事基因story-index.md— 主索引
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

