AI-Augmented Software Engineering 강의 시리즈의 과정 자료.

코스 개요 및 syllabus는 코스 웹 사이트에 있습니다.

이 저장소는 강의 노트, 슬라이드, 코드 예제, 할당 및 리소스를 수집하는 방법을 현대 AI 도구 변경 방법을 탐구하는 방법 소프트웨어 설계, 내장, 테스트, 유지 보수.

과정 개요

AI-a 증강 소프트웨어 엔지니어링은 AI 시스템 — 코드 조수, 에이전트 및 LLM 기반 도구 — 소프트웨어 개발 수명주기 전반에 걸쳐 활성 콜러레이터로서의 연습입니다. 이 과정은 how (이 도구의 실제 사용)과 why (강도, 한계 및 공학 연습을위한 임의)를 다룹니다.

강의

1. Setup Environments - Git, GitHub, Python 및 과정의 나머지 전에 코딩 에이전트를 설치합니다.
2. Specs-Driven 발달 — SDD 적합, 계획 형태, SDD 성숙 수준, 좋은 specs 및 SDD 기구가 있을 때 대리인 SDLC.
3. 코딩 에이전트의 Anatomy — 어떻게 LLMs 일, 모델 풍경, 왜 LLMs 코딩에 좋은, 그리고 후드 아래 코딩 에이전트.
4. Agentic Codebase 구조 - 기술, 규칙 및 난간; CLAUDE.md/AGENTS.md/SKILL.md/intent.md; 걸이; 그리고 시약 본.
5. 명세.

AI-Augmented Software Engineering

Course materials for the AI-Augmented Software Engineering lecture series.

The course outline and syllabus are on the course website.

This repository collects lecture notes, slides, code examples, assignments, and resources exploring how modern AI tools change the way software is designed, built, tested, and maintained.

Course Overview

AI-augmented software engineering is the practice of using AI systems — code assistants, agents, and LLM-based tooling — as active collaborators across the software development lifecycle. This course covers both the how (practical use of these tools) and the why (their strengths, limits, and implications for engineering practice).

Lectures

1. Setup Environments — install Git, GitHub, Python, and a coding agent before the rest of the course.
2. Specs-Driven Development — the agentic SDLC, when SDD fits, Plan Mode, SDD maturity levels, good specs, and SDD frameworks.
3. Anatomy of Coding Agents — how LLMs work, the model landscape, why LLMs are good at coding, and coding agents under the hood.
4. Agentic Codebase Structure — skills, rules, and guardrails; CLAUDE.md / AGENTS.md / SKILL.md / intent.md; hooks; and subagent patterns.
5. Advanced Context Management — a taxonomy of context, context assembly and discovery, the agent runtime pipeline, session management, and MCP fundamentals.
6. Agentic Code Review and Testing — deterministic guardrails, LLM-based review, safe deployment, runtime safety, and monitoring agents.
7. Agent Orchestration — the generator-critic loop, the goal pattern, the Ralph loop, running multiple agents, and the subagent pattern.
8. Security in the Age of AI — SAST/SCA and secret leaks, prompt injection and agent attack surfaces, and agent-assisted triage.
9. Software Factory and the Future of Software Engineering — self-running, self-improving systems and where AI software engineering goes next.
10. AI-Native Engineering Teams — how AI-native teams work, the standardized team stack, and LLM gateways, model routing, and cost optimization.

Midterm

The midterm (25 multiple-choice questions, 1 hour, 30% of the grade) covers Lectures 1–5 and their required readings. See the Midterm Study Guide.

Repository Structure

.
├── lectures/       # Lecture notes and slides, one folder per session
├── assignments/    # Homework and project briefs, one folder per lecture