paper-explainer
Skill46 starsChanged 5 months agoMITView on GitHub
AI论文解读工具。面向产品经理、运营等非技术背景但对AI感兴趣的读者,将晦涩的学术论文转化为图文并茂的通俗研报:讲清楚”这个技术解决什么问题、有多厉害、对业务意味着什么”,配以极简科技风插图,并输出莫兰迪风格的 HTML 单文件。
6 other projects use a version of this.
Adopted by
Most recently taken up by
- kakaxide/agent-skillstheir own version, first seen Nov 8, 2026
- ZJU-REAL/Easeltheir own version, first seen Oct 13, 2026
- yangwhale/CloseCrabtheir own version, first seen Aug 18, 2026
- sanqi-cd/Sanqi-Skillstheir own version, first seen Aug 3, 2026
- OvOhao/paper-explainertheir own version, first seen Jul 8, 2026
- idanbeck/claude-skillstheir own version, first seen Jan 19, 2026
Other versions6 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 6
- ZJU-REAL/Easel★ 4.5ksame name
- sanqi-cd/Sanqi-Skills★ 28same name
- idanbeck/claude-skills★ 14same name
- yangwhale/CloseCrab★ 5same name
- kakaxide/agent-skills★ 0same name
- OvOhao/paper-explainer★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
Files
Paper Explainer Skill
这是一个专门用于解读 AI/CS 领域技术论文的技能。
目标读者:产品经理、运营、业务负责人——对 AI 感兴趣,但没有算法/工程背景。
核心目标:去掉学术腔和工程术语,用”讲故事”的方式说清楚三件事:
- 这个技术在解决什么痛点?
- 它怎么做到的?(用类比,不用公式)
- 对我们的产品/业务意味着什么机会?
核心能力
- 故事化叙事:不复述摘要,而是从”旧方案有什么死穴”开始讲,讲到”新突破怎么绕过去的”。
- 视觉化解构:对最难懂的概念画”极简对比图”或”流程演进图”,让不懂技术的人看一眼就明白。
- 沉浸式研报:生成莫兰迪配色的 HTML 单文件,带关键数字卡片、结论高亮块,提供优质的阅读体验。
依赖说明
- 图像生成:本技能将直接生成 SVG 矢量图代码,无需外部 API 调用。这能确保生成的图表精准、清晰且风格统一。
- 联网搜索:需要调用 WebSearch 或 WebFetch 工具来处理 URL 输入。
工作流程 (Workflow)
当用户提供一篇论文(PDF、文本或 URL链接)并要求解读时,在开始之前先读取 prompts/narrative.md 和 prompts/visual.md,然后按以下步骤执行:
Step 0: 内容获取
- 本地文件:直接读取。
- URL:用
WebFetch抓取;内容截断时用WebSearch找 arXiv 纯文本页重新抓取;过滤导航/广告,只保留正文。
Step 1: 理解论文,构建叙事脉络
读完全文后,在内存中整理三个问题的答案:
- 旧方案的死穴是什么?(对应 Phase 1)
- 新方案的核心思路是什么?最适合用什么类比解释?(对应 Phase 2)
- 对 PM 而言最重要的结论和机会点是什么?(对应 Phase 4)
同时标记 2-3 个”视觉锚点”:最抽象、用文字说不清楚、但画一张图就能秒懂的概念。优先考虑 Before/After 对比图。
Step 2: 按叙事框架撰写正文
严格按照 prompts/narrative.md 的五个 Phase 撰写,写到视觉锚点时直接内联生成 SVG(参考 prompts/visual.md 的规范)。
SVG 生成要点:
- 先选图表类型(对比图/流程图/模块图),再动笔
- SVG 直接嵌入文档,不保存为外部文件
- 每张图只讲一个概念,不超过 5 个元素
Step 3: 生成 HTML 研报
按照 prompts/html_style.md 的规范将正文转为完整 HTML:
- Phase 0 之后必须插入
stat-cards,展示 1-3 个最重要的对比数字 - Phase 4 的”选型建议”必须用
insight-box呈现 - 代价/风险用
warning-box标出 - SVG 直接内联到 HTML 中,不使用
<img>外链 - 保存为
./output/{paper_keyword}_report.html
文件管理规范 (File Management)
- 输出根目录:
./output/(相对于当前项目工作目录) - 图片存储路径:
./output/assets/images/(自动创建) - 图片命名格式:
{paper_keyword}_{concept}.svg- 例如:
deepseek_r1_grpo.svg
- 例如:
- HTML 报告路径:
./output/reports/(建议新建此目录) - 提示词记录:(可选) 将生成的 prompt 保存到
./output/assets/prompts/以便复用。
资源索引 (相对于本技能文件夹)
- 叙事框架:
prompts/narrative.md - 绘图规范:
prompts/visual.md - HTML 样式:
prompts/html_style.md
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.

