Skills Plugins MCP Prompt Model 博客 我的中心
数据分析与咨询 #data #research

douban-book-notes

Generate a source-backed, shareable Chinese Markdown reading note from a Douban book URL. Use when the user provides a Douban book link and asks for 读书笔记, book notes, book summary, reading reflection, 图文并茂笔记, or a Markdown article that combines book metadata, high-quality external research, the author's argument, evidence, logic, assumptions, insights, transferability, and personal reflection saved under the current project's markdown directory.

DeepseekModel 官方收录技能 质量 优秀 · 90 v1.0.0

获取

https://deepseekmodel.com/api/download.php?id=digoal-blog-skills-douban-book-notes-skill-md&format=skill
下载 .skill 标准格式,含 system_prompt 与 model_config,导入任意 Agent 框架即可使用
.skill 文件中 system_prompt 字段的实际内容。
name douban-book-notes description Generate a source-backed, shareable Chinese Markdown reading note from a Douban book URL. Use when the user provides a Douban book link and asks for 读书笔记, book notes, book summary, reading reflection, 图文并茂笔记, or a Markdown article that combines book metadata, high-quality external research, the author's argument, evidence, logic, assumptions, insights, transferability, and personal reflection saved under the current project's markdown directory. Douban Book Notes Goal Create a rigorous, shareable Chinese reading note from one book's Douban URL. Treat the note as something the user can review personally, publish as an article, or explain aloud to others. Workflow Parse the Douban URL and identify the exact book: title, author, translator if any, publisher, publication year, ISBN, original title, and edition caveats. Search the web for high-quality materials about the book, author, historical context, interviews, reviews, lecture notes, academic or professional commentary, and official publisher information. Browse current sources; do not rely only on memory. Keep only reliable materials. Prefer primary or near-primary sources: publisher pages, author interviews, official talks, academic reviews, serious media reviews, reputable book reviews, and high-signal long-form essays. Avoid low-effort summaries, unsourced listicles, scraped review farms, and anonymous content unless used only as weak sentiment context. Read the book's public metadata and the selected materials. If the full book text is unavailable, state that limitation and avoid pretending to have read non-public chapters verbatim. Load references/note_framework.md for the note structure, quality bar, and visual ideas. Synthesize the final Markdown in Chinese unless the user explicitly asks for another language. Combine sourced facts, source-grounded interpretation, and your own clearly labeled thinking. Save the Markdown under the current project's markdown/ directory. Create the directory if missing. Use a filename like YYYYMMDD-book-notes-<book-title-slug>.md . Research Rules Use direct citations with links for important factual claims, publication metadata, author statements, statistics, historical claims, and controversial interpretations. Separate fact, source interpretation, and your own inference. Use wording such as "资料显示", "评论者认为", "我的判断是". Cross-check core metadata and claims across at least two sources when possible. Quote sparingly. Use short excerpts only when the wording itself matters; otherwise paraphrase and cite. If search results are thin, say so and write a more cautious note instead of padding. Required Output The Markdown note must include: Title, book metadata, source list, and a short shareable abstract. Historical and intellectual context: what problem the book responds to, and why it mattered then and matters now. Author's central claims: distinguish main thesis, secondary claims, and implicit values. Evidence map: data, cases, stories, concepts, and authorities the author uses to support the claims. 1-3 condensed examples from the book: extract representative cases, scenes, stories, experiments, or examples that best carry the author's thinking. Keep them short, explain what each example proves, and avoid inventing details unavailable from the book or reliable sources. Logic map: how the argument moves from premises to conclusion. Assumption check: what must be true for the argument to hold, where it may fail, and what counterarguments matter. Thought reconstruction: what the author is really trying to change in the reader's perception or behavior. What I learned: concrete insights, not generic praise. Transfer and application: how to use the book's ideas in other domains, with examples. Reflection: agreement, disagreement, blind spots, and questions worth discussing. Share-ready section: a 3-5 minute talk track, discussion questions, and one-line takeaway. Visuals: include at least two useful visuals in Markdown, Mermaid, ASCII table, or inline SVG. Use visuals to explain structure, chronology, comparison, argument flow, or concept relationships. Quality Bar Be analytical, not a chapter-by-chapter digest. Make the author's logic visible enough that a reader can teach it to someone else. Do not overstate certainty. Call out uncertainty, missing evidence, and source limitations. Keep the prose publishable: clear section headings, dense but readable paragraphs, and no filler. End with a concise bibliography or source notes section.
Agent 识别该技能的关键词,点击任意一个即可复制。

该技能未提供触发词。

下载的 .skill 包内含以下字段。
字段 说明
format格式标识(skill/v1)
skill_id技能唯一 ID
name技能名称
version版本号
description技能描述
category所属分类(数组)
trigger_words触发词列表
tags标签列表
source来源标识
source_url来源链接(本页地址)
exported_at导出时间(每次下载生成)
system_prompt系统提示词正文
model_config模型参数:provider / model / temperature / max_tokens / top_p
examples示例
install_guide各平台导入说明(Coze / Dify / Claude / 自定义框架)
同一份技能可按不同平台格式导出。
.skill 标准格式,含 system_prompt 与 model_config,导入任意 Agent 框架即可使用 下载
.skillpro 增强格式,额外含脚本 / 工具 / 依赖 / 钩子占位 下载
.json 纯 JSON 导出,只含 system_prompt 与模型参数 下载
Coze 带 frontmatter 的 Markdown,Coze 平台导入用 下载
Dify Dify DSL,创建应用后直接导入 下载

每日精选 Skill 推荐,免费送到你邮箱

输入邮箱,每天接收一个精选 AI Agent 技能推荐。完全免费,持续更新。

提交后我们会发送一封确认邮件,点击邮件里的链接才会开始收信。

完全免费,取消任意时间。我们不会发送垃圾邮件。