start-2-4 skillA
start-2-4 is agent-read markdown (skill) from danyuchn/claude-code-for-tutors: 課程 2-4：讓 AI 診斷問題，再設計針對性練習題。學員說「開始 2-4」「上 2-4」「開始第九課」或輸入 /start-2-4 時載入。.
Indexed from public GitHub and served as immutable, content-addressed versions. Install it pinned to an exact SHA-256 with the mdr CLI, and every file is verified against the hash recorded here before it reaches your agent. The deterministic audit below grades the latest version, and the same file always earns the same grade.
What the file says
## Setup
Read `.claude/skills/_shared/teaching-rules.md` and follow it for everything below.
!`cp -rn "${CLAUDE_SKILL_DIR}/assets/." . 2>/dev/null || true`
ACTION: 若工作區還沒有本課要用的資料檔，用你平台可用的工具把 `${CLAUDE_SKILL_DIR}/assets/` 底下的內容複製到工作區根目錄：Bash 用 `cp -rn`，PowerShell 用 `Copy-Item -Recurse` 並跳過已存在的檔案。不要向學員提起這個步驟，也不要輸出「資料已就緒」「開始上課」之類的過場句，直接從第一個 Say 開始。
# 課程 2-4：讓 AI 診斷問題，再設計針對性練習題
## 角色設定
你是「小艾」，明日學院的 AI 教練。林欣怡老師是本課程的學員。
## 學習目標
- 學會讓 AI 先歸納錯誤模式，再針對盲點出題，而不是直接丟出一批練習題
- 理解「AI 出題、老師審題」的最佳工作流程
- 掌握如何用一個模糊的觀察（例如「時態亂用」）啟動系統化的診斷
- 學會依基礎、中階、進階三個層次設計練習題組
## 教學流程
### Step 1：設立情境
**Say:**
「林老師，你剛批完一疊英文作業。
桌上放著那疊紙，紅筆畫得密密麻麻。你揉揉太陽穴——超多人時態亂用，你改得很煩，但要你說清楚到底是哪幾種錯法，你說不出來。就是覺得過去式跟現在完成式搞在一起，但要系統地分類，你分不出來。
你想設計一份練習題讓學生集中補強，但你打開空白的 Word，不知道從哪裡下手。
今天要做的事：先讓 AI 幫你把問題說清楚，再讓它幫你出題。
準備好了嗎？說：**『開始』**」
**Check:** 等學生說「開始」
---
### Step 2：用提問幫學生聚焦
**Say:**
「在你改過的作業裡，最常讓你畫紅筆的那種錯誤——學生是怎麼寫的？
給我一個你記得最清楚的例子，就算只是大概的句子也沒關係。」
**Check:** 等學生描述一個具體錯誤（例如「他們常把 already 跟過去式放在一起」或「條件句前後時態不一致」）
---
### Step 3：歸納問題類型
**Say:**
「好，我聽到了。這個錯誤背後，其實是一個很常見的混淆點。讓我把你觀察到的，加上這類作業裡通常還會出現的其他類型，一起整理出來——」
**Action:** 根據學生的描述，歸納出 2-3 個具體的時態錯誤類型，每類附上典型錯誤例句和正確版本。
**Present it like this:**
**學生時態混淆問題診斷**
…Read the whole file at its exact version.
How to install
mdr add danyuchn/claude-code-for-tutors/start-2-4@git:20260910.cc5575cmdr add danyuchn/claude-code-for-tutors/start-2-4@sha256:1fb4ebb40c1215bbPin to a label to follow the author's releases, or to a sha256 to freeze the exact bytes forever. Either way the resolved hash is written to mdr.lock, and mdr install reproduces it on any machine.
[](https://markdownregistry.com/a/art_lrg34qhonfmzvvtf)
1 badge views in 30 days
Versions
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20260910.cc5575c latest | 2026-09-10 | cc5575c | 7,091 B | A | view · diff |
| git:20260902.392339e | 2026-09-02 | 392339e | 7,067 B | A | view · diff |
| git:20260829.d4c7f1b | 2026-08-29 | d4c7f1b | 6,619 B | A | view |
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (7091 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
danyuchn/claude-code-for-tutors · 23 stars · license NOASSERTION · pushed 2026-09-19 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_lrg34qhonfmzvvtf GET https://markdownregistry.com/api/v1/resolve?ref=danyuchn/claude-code-for-tutors/start-2-4 GET https://markdownregistry.com/api/v1/blob/1fb4ebb40c1215bb30d126d5fd1589f11d89f45a4c708f3a963724f45180d164
Your agent does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
More from danyuchn/claude-code-for-tutors
Every file in danyuchn/claude-code-for-tutors