메뉴
Use when user wants to analyze dialogue style, extract speech patterns, compare character voices, or detect AI-sounding dialogue. Triggers include "对话", "口头禅", "角色声音", "对白", "AI味对话".
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Use when user wants to write novels, manage outlines, characters, world-building, style, workflows, or chapter review. Triggers: 写章节, 生成大纲, 创建角色, 世界观, 风格, 伏笔, 小说
Use when user wants to start or continue a long-session planning flow for a novel, gather and refine ideation, characters, setting, outline, or source packs, and hand off to Dante when the writing window is ready.
Use when user wants to write chapters, generate drafts, or continue the story. Triggers include "写第", "生成章节", "续写", "草稿", "创作".
Use when user wants to manage project settings, create characters, update world-building, view outline, or check foreshadowing status. Triggers include "新建", "角色", "世界观", "大纲", "伏笔", "项目".
Use when user wants to check logic consistency, polish style, find plot holes, or review existing chapters. Triggers include "检查", "润色", "逻辑", "伏笔漏洞", "审查".
Use when user wants to initialize style, extract reusable writing signals from user-supplied text, compose a project style guide, or analyze style drift.
| name | dialoguequality |
| description | Use when user wants to analyze dialogue style, extract speech patterns, compare character voices, or detect AI-sounding dialogue. Triggers include "对话", "口头禅", "角色声音", "对白", "AI味对话". |
用于分析角色对白、提取说话习惯、检查角色声音是否区分明显，以及识别 AI 味对话问题。
对话分析不要只盯着对白片段本身。当前推荐读取顺序是：
data/manuscript/arc_*/ch_*.md — 实际章节正文src/characters/*.md — 角色单源文档data/style/composed.md — 当前作品风格约束src/outline.md / canonical packet — 本章戏剧位置与章节目标如果需要完整判断某章对白是否失真，优先结合 review_chapter 或章节 packet，而不是只看几行台词。
| 工具 | 说明 |
|---|---|
extract_dialogue_fingerprint | 提取角色对白指纹 |
review_chapter | 对整章做审查，可发现台词失真、解释性对话、套路对白 |
先定位章节和角色，再提取：
用户: 分析一下林月在第六章的对话风格
AI:
[COMMAND] extract_dialogue_fingerprint {
"chapter_id": "ch_006",
"character_names": ["林月"]
}
重点观察：
avg_sentence_length：句长偏短还是偏长question_ratio：疑问句比例common_bigrams：高频词组speech_patterns：口头禅、重复习惯、语气特征提取完后，要对照 src/characters/*.md：
重点查这些问题：
data/style/composed.md对话检查结果：ch_006 / 林月
1. 角色声音
- 句长偏短，命令句比例高，符合技术组长的控制型表达
- 高频表达偏直接，几乎不解释理由，角色识别度高
2. 问题
- 第 3 段存在解释性对话，林月不该向熟悉项目的人完整解释背景
- 第 6 段礼貌度偏高，不像她一贯说话方式
3. 修改建议
- 删掉完整解释，改成一句结论 + 一个动作
- 把“请你尽快处理”改成“今天收掉”
❌
"你也知道，我们项目组最近一直在做那套老系统重构，所以今天才会出这种问题。"
✅
"老系统又炸了。"
原因：后者更像熟人之间的真实说话方式，信息由上下文承担，不靠台词硬讲。
❌ 陈明 / 林月 / 赵磊都用类似的完整句式解释问题
✅
陈明：偏技术比喻、偏自嘲
林月：短句、结论先行
赵磊：口语化、插科打诨
extract_dialogue_fingerprint 适合看“说话指纹”review_chapter 适合看“对白放在整章里是否成立”