課程目標與核心內涵
在生成式 AI 與大型語言模型（LLM）爆發的時代，軟體開發的門檻已被徹底顛覆。「Vibe Coding」指的正是開發者不再糾結於底層的語法細節，而是以高階導向、直覺溝通的「氛圍（Vibe）」來引導 AI 完成程式撰寫。

本課程核心理念為「有嘴萬事足」，旨在培養學生在 AI 時代身為「科技指揮官」的關鍵能力。

課程目標（Course Objectives）本課程期望透過密集的實作與觀念拆解，帶領學生達成以下四核心目標：破除程式焦慮，重塑開發思維： 讓無程式背景的學生克服對「寫 Code/抓 Bug」的恐懼，轉確認識「自然語言（人話）」也能成為高效開發工具。掌握 AI 協作精髓（Prompt Programming）： 學習如何精準定義需求、設定 AI 角色、提供脈絡，將模糊的想法轉化為 AI 能聽懂的結構化指令。具備「最小可行性產品（MVP）」的敏捷開發能力： 透過 Cursor、v0 等工具，在短時間內從零到一生成網頁、小工具或自動化流程，快速驗證自身創意。培養「導演思維」而非「代碼工匠」： 提升學生在數位專案中的架構設計、邏輯拆解與問題解決（Troubleshooting）能力。課程核心內涵（Course Connotation）課程內容圍繞三大核心維度展開，由淺入深，強調「即學、即說、即用」：1. 自然語言的「精準編譯」：把話說對，程式就對指令工程（Prompt Engineering）基礎： 學習如何像主管交代任務一樣，給 AI 清楚的脈絡（Context）、限制條件（Constraints）與期望輸出。邏輯拆解力： 練習將一個大專案（如：做一個記帳網頁）拆解成 AI 能逐步執行的微小單元。2. 前沿 AI 工具鏈的實戰應用：動口不動手原型一鍵生成： 利用 v0 by Vercel，用嘴說出介面想像，幾秒內生成現代化的前端 UI 介面。AI 原生編輯器實作： 深入體驗 Cursor，學會如何使用語音輸入、@ 檔案引用與行內編輯（Inline Edit），讓 AI 自動修復 Bug、擴充功能。3. 雙向反饋與架構優化：當個稱職的「程式碼總監」程式碼的「審查（Code Review）」能力： 雖然不用親自寫，但要看得懂結構。學習如何引導 AI 解釋程式碼，並在出錯時透過對話引導 AI 自我修正。人機協作的 Vibe 養成： 體會何時該讓 AI 自由發揮（Creative Generation），何時該強加約束（Strict Architecture）。課程亮點 這不是一堂程式語言課，而是一堂「超能力溝通課」。學生將在課堂上體驗到：過去需要學海無涯的 JavaScript/Python 才能做出的專案，現在只要「嘴力全開」，就能在 2 小時內看見成品在網頁上運作。這是一場解放創意、讓每個人都能成為 Full-Stack 創作者的啟蒙實踐。

Course objectives and core connotations
In the era of the explosion of generative AI and large language models (LLM), the threshold of software development has been completely subverted. "Vibe Coding" refers to the fact that developers no longer dwell on the underlying syntax details, but use a high-level oriented, intuitive communication "vibe" to guide AI to complete program writing.

The core concept of this course is "If you have a mouth, you can do everything", it aims to cultivate students' key abilities as "technological commanders" in the AI ​​era.

Course Objectives: This course hopes to lead students to achieve the following four core goals through intensive practice and concept dismantling: to eliminate programming anxiety and reshape development thinking: to enable students without programming background to overcome the fear of "writing code/catching bugs" and to transform their understanding that "natural language (human language)" can also become an efficient development tool. Master the essence of AI collaboration (Prompt Programming): Learn how to accurately define requirements, set AI roles, provide context, and transform vague ideas into structured instructions that AI can understand. Agile development capabilities with "Minimum Viable Product (MVP)": Use tools such as Cursor and v0 to generate web pages, gadgets or automated processes from scratch in a short time to quickly verify your own ideas. Cultivate "Director's Thinking" instead of "Code Craftsman": Improve students' architectural design, logical disassembly and problem-solving (Troubleshooting) abilities in digital projects. Course Connotation The course content revolves around three core dimensions, from the shallower to the deeper, emphasizing "ready-to-learn, ready-to-speak, ready-to-use": 1. "Accurate compilation" of natural language: if you say the words correctly, the program will be the basis of command engineering (Prompt Engineering): Learn how to give AI a clear context (Context), constraints (Constraints) and expected output like a supervisor handing over a task. Logical disassembly power: Practice breaking down a large project (such as making an accounting web page) into tiny units that AI can execute step by step. 2. Practical application of cutting-edge AI tool chain: One-click prototype generation without using hands-on words: Use v0 by Vercel to speak your interface imagination and generate a modern front-end UI interface in a few seconds. AI native editor implementation: Experience Cursor in depth, learn how to use voice input, @ file references and inline editing (Inline Edit), and let AI automatically fix bugs and expand functions. 3. Two-way feedback and structure optimization: Be a competent "code director" and have the "Code Review" ability of the code: Although you don't have to write it yourself, you need to understand the structure. Learn how to guide the AI ​​to interpret code and use dialogue to guide the AI ​​to correct itself when errors occur. Vibe development of human-machine collaboration: Understand when to let AI play freely (Creative Generation) and when to impose constraints (Strict Architecture). Course Highlights This is not a programming language class, but a "superpower communication class." Students will experience in class: projects that used to require endless knowledge of JavaScript/Python can now be seen running on the web page within 2 hours as long as they "open their mouths". This is an enlightenment practice that liberates creativity and allows everyone to become a Full-Stack creator.

自編教材

Self-edited teaching materials