Software Development with AI Coding Agents

This course is a hands-on introduction to software development using AI coding agents. It focuses on using agents like Claude Code to automate workflows, ensure code quality, and increase developer productivity.

Labs/Exercises

There are exercises available for all topics covered.

Agent

This course is taught with Claude Code and Claude CLI by default. But it can be taught with any programming language, development environment and agentic agents

Prerequisites

Participants need to have access to the coding agent used in the course (Claude Code by default).

1: Introduction

• Course Introduction
• Group Introduction
• Intro to AI prompting
• Intro to Agentic AI
• Understanding LLMs: Why they work and where they fail
• Managing Context: Preventing ‘hallucinations’ and keeping the agent focused
• Prompting vs. Agentic Coding: Moving from one-off questions to autonomous workflows
• Why and when to use agentic AI
• Ethical, security, and privacy concerns

2: Effective Prompting & Workflows

• Effective prompting
• Prompting for common development tasks
  • Accelerating code generation
  • Safe refactoring at scale
  • Writing meaningful documentation and comments
  • Automated unit test generation
• Avoiding common pitfalls
• Configuration
  • Project-specific memory (e.g. claude.md)
  • Defining project rules and standards

3: Mastering Agentic AI

• Best practices: ensuring quality and correctness
• Verify results using Spec-driven development
• Writing robust specifications
• Optimizing for Token Efficiency
  • Reducing latency and cost
  • Using Caching and Compaction effectively

4: Advanced Features & Ecosystem

• Extending the agent’s capabilities
  • Custom skills, commands and tools
  • Orchestrating subagents for complex tasks
  • Model Context Protocol (MCP): Connecting agents to your data and tools