AI-Assisted Development

Modified

May 19, 2026

Software development is changing faster than any framework or language ever has. AI coding assistants have moved from autocomplete curiosities to agents capable of planning, writing, and reviewing entire features autonomously. This module guides you how to work with that shift rather than around it, covering the mental models, tools, and workflows that let developers stay in control while letting AI handle the repetitive, mechanical, and context-switching parts of the job.

Learning Objectives

By the end of this module, you will be able to:

  • Apply the three stages of AI-assisted development (vibe coding, agentic development, spec-driven development) to real projects.
  • Configure and use AI coding tools such as GitHub Copilot, MCP servers, and agent skills.
  • Design effective prompts and specifications that keep you in control of the generated output.
  • Evaluate when to delegate work to an AI agent and when to write code by hand.

Session Outline

# Session Topics Covered
1 Vibe Coding First stage of AI-assisted development
2 Spec-Driven Development Spec-first workflow, planning, tasks
3 Get Started with SDD Hands-on Spec Kit setup
4 Prompt Engineering Designing effective prompts
5 Skills Reusable agent skills
6 MCP Model Context Protocol
7 GitHub Copilot Copilot in practice