|

课后实操指南
把第一节课变成你自己的能力
Post-Class Practice Guide
Turn Your First Class Into Real Skill

恭喜你上完第一节课!但 听懂 ≠ 会用。真正的能力,来自你亲手在 cowork 里走一遍。 下面五步,请按顺序做,每一步都别跳过「记录心得」——这是整个方法的灵魂。Congrats on finishing Session 1! But understanding ≠ being able to do it. Real skill comes from doing it yourself in Cowork. Follow these five steps in order. Don't skip the "write reflections" part — it's the soul of this method.

材料从哪来Materials First 先下载,再开始— Download Before You Begin

开始前,先回到 S1 学员入口页(你点进来的那个页面),下载两样东西:Go back to the S1 Student Portal page (where you came from) and download two things:

  1. 课程材料包Course Materials Package——点「下载课程材料(含作业)」按钮,下载后 解压。里面有 跟我做-提示词.mdstarter-project/、两个 JD,还有 homework-guide.md这份就是作业要求),下面都会用到。Click "Download Course Materials (includes homework)" and unzip it. It contains 跟我做-提示词.md, starter-project/, two JDs, and homework-guide.md (this is your homework requirements) — all of which you'll use below.
  2. 课程 PPTCourse Slides——点「下载 Session 1 幻灯片(PPTX)」按钮。Click "Download Session 1 Slides (PPTX)".
💡 下面第 0 步要放进 ppt/ 的 PPT、要放进 课程心得/跟我做-提示词.md,以及要放进 作业项目/ 的作业要求(就是解压包里的 homework-guide.md,放进去后改名 作业要求.md),都来自这两个下载。
另外:跟着 跟我做-提示词.md 跑课堂 demo 时,会让你新建 另一个 指向 starter-project 的 Cowork Project——那是独立 Project,和这里的「我的AI课程」并存、随时切换、互不影响。
In Step 0 below, the PPT goes into ppt/, 跟我做-提示词.md goes into 课程心得/, and the homework requirements (the homework-guide.md from the unzipped package, renamed to 作业要求.md) all come from these two downloads.
Also: when following 跟我做-提示词.md for the class demo, you'll create another separate Cowork Project pointing to starter-project — it coexists with your "MyAICourse" project, you can switch between them anytime, and they don't affect each other.

第 0 步Step 0 · 先搭好你的学习工作区Set Up Your Learning Workspace

打开 cowork,新建一个 空目录 作为你的课程学习项目(比如 我的AI课程)。 空的就对了——接下来的内容由你和 AI 一起长出来。Open Cowork and create an empty directory as your course project (e.g. 我的AI课程). Starting empty is correct — you and the AI will grow the content together.

建议的目录结构(边做边建,不用一次到位):Suggested directory structure (build it gradually, not all at once):

我的AI课程/
├── 课程心得/
│   ├── ppt/                  # 把课程 PPT 放这里
│   ├── 跟我做-提示词.md       # 下载包里的演示提示词清单,每步带「这步在干嘛」(很有用!)
│   ├── 我的心得.md            # 随手记,想到啥写啥
│   ├── 不懂的名词.md          # 听到没懂的术语/概念
│   └── 课程总结.md            # 让 AI 帮你生成
├── 课堂项目/                  # 跟着课程自己做一遍
└── 作业项目/
    └── 作业要求.md            # 下载包里的 homework-guide.md 放这,改名即可
💡 .md?用 .txt 一样行。Scared of .md? .txt works just fine. 上面那些 .md 文件,你嫌麻烦就全建成 .txt(系统自带记事本就能写), AI 照样读得懂、照样帮你总结。.md 只是排版能好看一点,不影响用——别让格式挡住你动手。All those .md files can be created as .txt instead (any built-in text editor works), and AI can read and summarize them just the same. .md just looks nicer when formatted — don't let file format stop you from getting started.
💬 真实示例(直接抄)Real Example (copy directly)——不用自己一个个手动建文件夹,把上面那段结构整段丢给 AI:No need to manually create each folder — just paste the entire structure above to AI:

「帮我建下面的目录结构,文件我会慢慢放」
(把上面「我的AI课程/…」目录结构整段粘进去)
"Help me create the following directory structure, I'll add files gradually"
(paste the entire 「我的AI课程/…」 structure above)

这步在干嘛:让 AI 一次性把整个课程文件夹骨架建好,你只管之后往里放文件。What this does: lets AI build the entire course folder skeleton at once, and you just add files as you go.

第 1 步Step 1 · 记心得 + 让 AI 帮你总结Write Reflections + Let AI Summarize

先你写,再让 AI 补。Write first, then let AI fill in. 顺序很重要——AI 是来帮你深化的,不是替你回忆的。Order matters — AI is here to deepen your thinking, not recall it for you.

  1. 我的心得.md 里,尽量回忆课上听到的内容,记下你的体会。不用写得完善,碎片、半句话都行。In 我的心得.md, recall what you heard in class and write down your thoughts. No need to be thorough — fragments and half-sentences are fine.
  2. 不懂的名词.md 里,记下你没太懂的术语和概念。In 不懂的名词.md, note down any terms and concepts you didn't quite understand.
  3. 把 PPT 放进 ppt/ 目录;再把下载包里的 跟我做-提示词.md 也放进 课程心得/——它是课堂 demo 用到的 提示词清单,每步还标了「这步在干嘛」,AI 翻它能准确还原每步在做什么。Put the PPT in the ppt/ directory; also put 跟我做-提示词.md from the download package into 课程心得/ — it's the prompt list used in the class demo, with each step annotated with "what this step does", so AI can accurately reconstruct what each step is doing.
  4. 然后让 AI 结合你的心得、PPT 和 跟我做-提示词.md 帮你总结。多问几次,每次问得更细,最后要求它 输出一个 课程总结.md 文件,方便你以后回看。Then ask AI to summarize by combining your reflections, the PPT, and 跟我做-提示词.md. Ask multiple times, going deeper each time, and finally ask it to output a 课程总结.md file for future reference.
📎 跟我做-提示词.md 为什么重要Why 跟我做-提示词.md mattersPPT 是「讲了什么」,跟我做-提示词.md 是「实际发了哪些提示词」。 你对某一步没看懂、或想不起 demo 细节时,让 AI 翻 跟我做-提示词.md 回答你,比凭空解释准得多。The PPT covers "what was taught"; 跟我做-提示词.md covers "what prompts were actually sent". When you don't understand a step or can't remember demo details, asking AI to reference 跟我做-提示词.md gives much more accurate answers than explaining from scratch.
Prompt 示例Prompt Examples(直接抄,再改成你自己的):(copy directly, then adapt to your own):
💡 技巧Tip把「不懂的名词」逐个丢给 AI,让它用大白话 + 例子解释,再让它写回 不懂的名词.md。 一节课下来,你就有了一份属于自己的术语词典。Give each "unknown term" to AI one by one, ask it to explain in plain language with examples, then have it write back to 不懂的名词.md. By the end of the class, you'll have your own personal glossary.
💬 真实示例(一位同学真实输入的,直接抄、照着改)Real Example (from an actual student, copy and adapt)——核心是你只管 口述,开文件 / 写入 / 排版全交给 AI:The key is you just dictate, and AI handles opening files / writing / formatting: 一句话窍门:你是指挥者——口述内容,建文件、写入、排版都交给 AI。One-line tip: you're the commander — dictate the content, and let AI handle creating files, writing, and formatting.

第 2 步Step 2 · 亲手做课堂项目Do the Class Project Yourself

只有自己做,才有真体会。You only truly learn by doing. 看录像、对着课程资料,自己在 课堂项目/ 里从头做一遍。Watch the recording, follow the materials, and redo everything from scratch in the 课堂项目/ folder.

第 3 步Step 3 · 做作业 = 解决你自己的真实痛点Homework = Solve Your Real Problem

作业不是交差,是 用这节课的能力解决你日常工作里的一个痛点。项目可大可小。Homework isn't just a formality — it's about using this class's skills to solve a real pain point in your daily work. The project can be large or small.

先想痛点Start by identifying your pain point——它可能是:it might be:

然后分两种情况:Then handle it in one of two ways:

你的情况Your situation怎么做What to do
痛点 已经有固定流程Pain point already has a fixed process试着把这个流程搬进 cowork,做成一个 agent / workflowTry moving that process into Cowork and turning it into an agent / workflow
痛点 还没成型Pain point doesn't have a clear process yet把「作业要求 + 你的痛点描述」一起给 AI,让它给你建议方案,保存成文件Give AI both "homework requirements + your pain point description" and ask it for a suggested solution, saved as a file

关键是迭代The key is iteration——拿到 AI 的建议后,别停。和它一起细化,就像我做 agent 时不断完善 prompt 那样:你提一点、它改一版、你再提……几轮下来,方案就成型了。全程继续写你的心得体会。Once you have AI's suggestions, don't stop there. Refine together, just like I kept improving prompts when building agents: you suggest something, it revises, you suggest again… after a few rounds, the solution takes shape. Keep writing your reflections throughout.

Prompt 示例Prompt Example
这是我的作业要求(作业要求.md),我日常工作的痛点是 ___。请帮我分析这个痛点适不适合用 AI/cowork 解决,给我 2-3 个方案,保存成 方案.md,并告诉我最简单能先跑起来的是哪个。」Here are my homework requirements (作业要求.md). My daily work pain point is ___. Please help me analyze whether this pain point is suitable for solving with AI/Cowork, give me 2-3 approaches, save them as 方案.md, and tell me which one is the simplest to get running first."
💬 真实示例(填好的样子,照着换成你的痛点)Real Example (filled in — swap your own pain point)
这是我的作业要求(作业要求.md,就是 homework-guide.md),我日常工作的痛点是 要花很多时间写周报。请帮我分析这个痛点适不适合用 AI/cowork 解决,给我 2-3 个方案,保存成 方案.md,并告诉我最简单能先跑起来的是哪个。」Here are my homework requirements (作业要求.md, i.e. homework-guide.md). My daily work pain point is spending too much time writing weekly reports. Please help me analyze whether this pain point is suitable for solving with AI/Cowork, give me 2-3 approaches, save them as 方案.md, and tell me which one is the simplest to get running first."

第 4 步Step 4 · 让 AI 帮你给我写一封邮件 = 提交作业Let AI Help You Write a Submission Email

最后,结合你这一路的 心得体会课后作业,让 AI 帮你写一封邮件发给我。Finally, draw on your reflections and homework throughout this process, and ask AI to help you write an email to send me.

📌 关键说明Important note这封邮件 就是你的作业提交——「作业要求」邮件里要的每一项内容,邮件里 一项都不能少。 让 AI 帮你写,不等于让 AI 帮你省。所以 prompt 里要把作业要求的每一项都点给 AI,让它逐项填满。This email is your homework submission — every item the "homework requirements" asks for must appear in the email, not a single one left out. Having AI help you write it doesn't mean having AI help you skip things. So make sure your prompt points out every requirement item so AI fills each one in.

作业要求里要交的(邮件正文必须全部覆盖):What the homework requires (the email body must cover all of these):

  1. 工作流设计Workflow design——你设计了哪几个 Agent,各自职责、数据怎么从上一个传到下一个。which Agents you designed, their respective responsibilities, and how data flows from one to the next.
  2. 每个 Agent 的提示词Each Agent's prompts——角色 / 任务 / 输出格式三要素齐全,并写明输入格式和 OUTPUTS/ 输出路径。with all three elements complete: role / task / output format, plus input format and OUTPUTS/ path.
  3. 实际运行输出Actual run output——每个 Agent 真实跑出来的结果(不是占位符、不是「[略]」)。the real output from each Agent (not placeholders, not "[omitted]").
  4. 学习心得Learning reflections——≥200 字,且至少 3 条具体的个人观察(「这工具好用」不算,要写你真发现了什么)。≥200 words, with at least 3 specific personal observations ("this tool is useful" doesn't count — write what you actually discovered).
Prompt 示例Prompt Example(直接抄,把括号里的路径换成你自己的):(copy directly, replace the paths in brackets with your own):
请结合我的作业项目(作业项目/ 目录)和课程心得(课程心得/ 目录),帮我写一封发给老师的作业提交邮件。作业要求的每一项内容都必须完整出现在邮件正文里,一项都不能省,按下面四块组织:Please combine my homework project (作业项目/ directory) and course reflections (课程心得/ directory) to help me write a homework submission email to the instructor. Every item the homework requires must appear completely in the email body — not one can be skipped. Organize it in these four sections:
  1. 工作流设计:列出我设计的每个 Agent、各自职责、Agent 之间怎么传数据;Workflow design: list each Agent I designed, their responsibilities, and how data passes between Agents;
  2. 每个 Agent 的提示词:原样贴出,确认角色/任务/输出格式三要素齐全,并带上输入格式和 OUTPUTS/ 路径;Each Agent's prompts: paste them as-is, confirming all three elements (role/task/output format) are complete, along with input format and OUTPUTS/ path;
  3. 实际运行输出:把每个 Agent 真实跑出的结果贴进来,不要用占位符;Actual run output: paste the real results from each Agent, no placeholders;
  4. 学习心得:≥200 字,至少 3 条我自己的具体观察。Learning reflections: ≥200 words, at least 3 of my own specific observations.
末尾再补一段我还想请教的困惑。语气真诚、简洁。写完先列一个清单核对上面四项有没有遗漏。」Add a paragraph at the end about questions I still have. Keep the tone sincere and concise. After writing, list a checklist to verify none of the four items above are missing."

我想看到的不是完美,而是 你真的动手了、真的在想。困惑和卡点尤其欢迎——那正是下一节课我能帮你的地方。What I want to see is not perfection, but that you truly worked on it and truly thought about it. Confusion and roadblocks are especially welcome — those are exactly where I can help you in the next class.

一句话总结One-sentence summary
写心得 → 让 AI 总结 → 自己做 → 解决真痛点 → 反馈给我。Write reflections → Let AI summarize → Do it yourself → Solve a real pain point → Send it back.
每一步都亲手做,每一步都留下记录。这门课的能力,是「做」出来的,不是「听」出来的。Do every step yourself, and document every step. The skill in this course is built by doing, not listening.

—— 看不懂任何一步,截图发学习群,老师和同学帮你。没有笨问题。—— If anything is unclear, screenshot it and post in the student group — the instructor and classmates will help. No question is too basic.