Module 1 of 3

What Coding Agents Are Doing

Cursor, GitHub Copilot, Claude Code, and Windsurf, and why this generation is different

The current generation of coding agents differs structurally from the inline-completion tools that preceded them. This module explains the shift from completion to agentic loops, what happens when a coding agent has access to a codebase and a shell, and why that changes the risk and value profile.

An Intro to Coding Agents
  • ~50 mins
  • 3 lessons
  • professional
Get this course

What you'll learn

  • Explain the structural difference between autocomplete-era coding tools and current agentic coding tools
  • Describe what a coding agent can access and act on in a typical engineering environment
  • Distinguish between the major coding agent products by their architectural capabilities and constraints
  • Articulate why the shift to agentic coding changes the risk profile for engineering organisations

Lessons in the Module

1 From Autocomplete to Agent 20–25 mins
2 What the Model Sees 20–25 mins
3 The Major Tools and Their Architecture 20–25 mins
Back to course overview