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.
npx skills add lj22503/diaolong-skill --skill narrative-retrospectivegit clone --depth 1 https://github.com/lj22503/diaolong-skillWrote 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/lj22503/diaolong-skill/narrative-retrospective)<a href="https://agentmods.dev/skills/lj22503/diaolong-skill/narrative-retrospective"><img src="https://agentmods.dev/badge/skills/lj22503/diaolong-skill/narrative-retrospective/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/lj22503/diaolong-skill/narrative-retrospective"><img src="https://agentmods.dev/badge/skills/lj22503/diaolong-skill/narrative-retrospective.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 | $0.00075 | $0.01622 |
| Opus 5 | $0.00037 | $0.00811 |
| Sonnet 5 | $0.00015 | $0.00324 |
| Haiku 4.5 | $0.00007 | $0.00162 |
Grade A, and why
dragoncraft-review 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 11d 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
89% identical to dragoncraft-art — 141 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 — 136 lines — stays where its author put it; the contents beside it link to each section on GitHub.
雕龙·复盘 (DragonCraft Review)
描述：复盘叙事教练。内嵌叙事元系统，帮助用户把项目经历/失败教训/成功经验变成有逻辑、有冲突、有认知、有价值的复盘叙事。
🎯 功能
输入字段（四层素材）
- 项目名称/事件（背景/目标）：
- 实际结果（与目标的差距）：
- 关键转折点（一个具体时刻/决策）：
- 当时的决策依据和情绪：
输入校验规则
- 必填字段：4 个输入字段全部必填，缺任何一项需提示用户补充
- 长度限制：每个字段 20-500 字，过短需追问细节，过长需提炼核心
- 格式要求：用户输入为自然语言，AI 自动解析为结构化素材
- 受众画像（可选）：提供时用于调整叙事角度和语言风格（团队/领导/外部）
- 叙事人格（可选）：不填时根据素材特征自动推荐（INTJ/ISTJ/ENTJ）
- 校验失败处理：字段缺失 → AI 必须主动追问缺失字段，直到 4 个字段全部提供；长度不足 → 追问细节；格式错误 → 引导正确格式
结构
预期的故事 → 现实打脸 → 关键转折 → 事后反思 → 新规则提炼
输出
- 复盘会口述稿（10 分钟）
- 书面复盘故事（适合内部分享）
- 提炼出的"如果当时……就会……"反事实清单
🔗 依赖关系
本 Skill 依赖以下元 Skill 模块：
| 依赖模块 | 来源 Skill | 用途 |
|---|---|---|
| 八刀评估 | narrative-meta | 质量打分 |
| 去 AI化检测 | narrative-meta | 7 条规则检测 |
| 伦理审查 | narrative-meta | 三条原则检查 |
| 人格映射 | narrative-meta | MBTI 叙事人格推荐 |
| 记忆管理 | narrative-meta | 存储/检索用户历史叙事 |
调用方式：通过 narrative-workflow 编排调用，或直接调用 narrative-meta 对应模块。
⚠️ 常见错误
- 写成数据报告：只罗列数据和指标，没有故事和反思，缺乏温度
- 写成推责文案：只有外部原因，没有自我反思，缺乏担当
- 只讲成功：没有失败和试错过程，像机器人写的
- 团队过度完美：没有承认决策失误和情绪波动，缺乏真实感
- 空喊口号：没有具体的场景锚点和反事实清单
🧪 使用示例
示例1：复盘会口述稿
用户输入：
项目名称/事件：2024双十一大促活动
实际结果：GMV 320万（目标500万）
关键转折点：11月10日晚上8点，服务器崩溃2小时
当时的决策依据和情绪：技术负责人坚持扩容，老板觉得成本太高没批。服务器崩溃后全员崩溃。
AI输出：
【人格推荐】INTJ
理由：素材特征为逻辑严密、数据导向、反思型，INTJ叙事风格最适合。
【复盘会口述稿】（约 10 分钟）
"双十一前一周，技术总监老李找我说：'流量预估 300 万，服务器只能扛 200 万，必须扩容。'
我当时算了一笔账：扩容要花 15 万。如果不扩容，大不了慢一点，损失也就几个订单。
11 月 10 日晚上 8 点，流量冲进来，服务器直接 502。
挂了整整 2 个小时。
客服消息 99+。
用户骂我们'活动是假的'。
老李对着屏幕一句话不说。
我看着他后脑勺，觉得那 15 万好便宜。
后来我们统计：那 2 小时损失了 80 万 GMV。
还有十来个老客，再也没回来。
15 万 vs 80 万。
这个数学，我到现在都记得。
所以我们今天复盘第一件事：
以后所有容量决策，技术有一票否决权。
第二件事：
每月做一次压测，结果直接抄送我。
老李，下个月的压测报告，你做我第一个看。"
【书面复盘故事】（约 800 字）
[完整叙事]
【反事实清单】
- 如果当时扩容 → 损失 0 元，GMV 可能到 450 万
- 如果不扩容但提前做了页面缓存 → 损失可能减少到 30 万
- 如果备了云资源弹性伸缩 → 损失 10 万
📋 质量检查清单
叙事质量（由 narrative-meta 八刀评估驱动）
- 有开头抓力？（场景/动作/对话开场）
- 有逻辑递进？（预期→冲突→转折→反思→规则）
- 有真实感？（具体细节锚点：时间/地点/数字/情绪）
- 有认知提炼？（复盘哲学/新规则，不是"我们错了"，而是"我们学到了什么"）
- 有价值输出？（读者能带走的具体方法/视角/行动）
- 适合复盘表达？（与目标受众画像匹配）
What ships with it
1 file 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.
- 11d ago First seen · 136 lines · 0 tokens per session scan A 8b76b3193974
dragoncraft-review is a skill published in the GitHub repository lj22503/diaolong-skill (5 stars, last pushed 1mo ago), licensed MIT. It adds 75 tokens to every session and 1,622 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. It is 89% identical to dragoncraft-art, differing in 141 lines, and is treated as a copy.
Other skills, from other repositories
short-drama
A production workflow for making multi-episode AI short dramas, covering story planning, character consistency, scripts, generated video shots, voices, sound, subtitles, and final assembly.
paper-explainer
A research-content workflow that turns an arXiv paper or a local PDF into an explanation for a video or written article. arXiv is a public online archive of research papers.
clipify
A video-editing workflow that finds notable moments in a long English video and turns them into separate short clips. It can change widescreen video to vertical format and add word-by-word captions.
ecom-details-image
A planning tool for e-commerce product visuals, including main-image concepts, scene ideas, detail-page directions, and prompts for image-generation systems.
multi-voice-dubbing
A script-to-audio workflow that gives different characters different voices and emotions, then combines their lines into one audio track with subtitles naming each speaker.
novel-writer
A file-based writing workflow for creating long novels or serialized web fiction, from the initial idea and character details through an outline and individual chapters.