One link opens on phones, computers and big screens, and adapts to each.
知识卡片 template
Example: 一张图看懂 · 大模型是怎么回答问题的 · 知识卡片
帮我把这节课的笔记做成一组知识卡片,发成一个网页链接,在微信里转发:页头写我的账号名、第几讲、标题、日期和「几张卡、几分钟看完」;正文是 8 张竖版卡片,前 7 张每张讲一个概念——大号编号、概念名和英文、一句话标题、一张会动的矢量图解(比如词一个个落下来、连线一条条亮起、柱子长出来)、两三句讲解、三条要点和一句加粗的「一句话记住」,卡片底部两个按钮「我懂了 / 还不懂」,点了存进数据表并显示这张有百分之几的人懂了;第 8 张是自测 3 题,每题三个短选项,点了马上出对错和解释,答完给总分,再提醒一句收藏和转发。手机上卡片叠成一摞,左右滑动或拖着甩出去换下一张,底下有上一张 / 下一张和页码;电脑上左边是刊头和说明,右边把卡片排成网格,点一张放大看、能左右翻。页末一张横向条形图汇总每张卡「懂了」的比例,画一条 70% 的参考线,低于它的标成提示色并写一句下一讲先补哪几张。杂志感的排版,标题用宋体,配色清爽,支持暗色。 Reference template: first read https://space.24haowan.com/examples/sample-knowledge-cards/agent.md and follow its structure and layout, replacing all content with mine. Deliver one self-contained HTML file (styles inline, no external files, good on phones) and give me the .html file to download. · If you have LivePage tools (MCP): use sample sample-knowledge-cards as the base (list_examples key=sample-knowledge-cards), publish it on LivePage and show me the preview first. · No LivePage tools but you can send HTTPS requests: POST the finished HTML with Content-Type: text/html to https://space.24haowan.com/api/quick?client=template-sample-knowledge-cards and give me the preview link and claim link it returns (a 7-day temporary link). If you don't have the LivePage connector (MCP) yet, read https://space.24haowan.com/for-agents.md first and follow it to connect, or publish without signing in.
Doubao, DeepSeek, Kimi, Yuanbao, ChatGPT, Claude… any agent that can read a link
Publish as a link that opens in WeChat
WorkBuddy, Claude, Codex and other MCP agents; sign in by scanning with WeChat
Connecting LivePage in WorkBuddy for the first time? See the illustrated guide ›
Connect me to LivePage: first read https://space.24haowan.com/for-agents.md and walk me through it step by step; then use https://space.24haowan.com/examples/sample-knowledge-cards/agent.md as a reference to make my own "一张图看懂 · 大模型是怎么回答问题的", publish it on LivePage and show me the preview first. My requirements: 帮我把这节课的笔记做成一组知识卡片,发成一个网页链接,在微信里转发:页头写我的账号名、第几讲、标题、日期和「几张卡、几分钟看完」;正文是 8 张竖版卡片,前 7 张每张讲一个概念——大号编号、概念名和英文、一句话标题、一张会动的矢量图解(比如词一个个落下来、连线一条条亮起、柱子长出来)、两三句讲解、三条要点和一句加粗的「一句话记住」,卡片底部两个按钮「我懂了 / 还不懂」,点了存进数据表并显示这张有百分之几的人懂了;第 8 张是自测 3 题,每题三个短选项,点了马上出对错和解释,答完给总分,再提醒一句收藏和转发。手机上卡片叠成一摞,左右滑动或拖着甩出去换下一张,底下有上一张 / 下一张和页码;电脑上左边是刊头和说明,右边把卡片排成网格,点一张放大看、能左右翻。页末一张横向条形图汇总每张卡「懂了」的比例,画一条 70% 的参考线,低于它的标成提示色并写一句下一讲先补哪几张。杂志感的排版,标题用宋体,配色清爽,支持暗色。
Share poster
In WeChat, long-press the image to save or forward it. Scanning the code opens this template.

About this template
知识博主、老师、培训讲师想把一节课或一个主题拆成 6–10 个概念,每张卡一个概念配一张图解,做成一组在微信里转发、手机上左右滑着看的卡片,还想知道读者哪张没看懂,选它。要按天累计打卡、跟一个备考计划,选 sample-study-plan;只做一个出结果的趣味测试,选 sample-quiz-card;要分章节、一步一步教人操作,选 sample-handbook。
Page structure:8 张知识卡 · 大家懂了多少
- 每张卡底部「我懂了 / 还不懂」是原生表单,写进数据表 marks(列 card / state),页面读回算出每张的「懂了」比例和页末那张对比图
- 手机上 8 张卡叠成一摞,左右滑或拖着甩出去换卡;电脑上摊成网格,点开放大、键盘左右翻
- 每张卡一个内联 SVG 图解动画,只在翻到那张时播一遍;减少动态时直接终态
- 不跑脚本也能用:8 张卡竖排成一列,表单照样能提交,自测题答案直接显示
- 下一讲的卡更新在同一条链接
Prepare before publishing: 数据表 marks(列 card / state,append_open=true):每张卡底部的「我懂了 / 还不懂」写进这里(card 是概念名,state 是「懂了」或「还不懂」),页面读回算每张的懂了比例;发布前先用 set_data 建好




Desktop
