Back to catalog
By CourseMade
LLM Coding Agents in Practice: Claude Code, Codex, and GPT
19 lessons4h
A hands-on, under-the-hood course for developers moving from chat-based AI to terminal-native agent tools. Learn how Claude Code, Codex, and GPT-based assistants manage prompts, context, and tool use, and master prompting, agent workflows, code review, security guardrails, and Git/CI-CD integration.
What you'll learn
Lessons unlock once the course is in your library.
Module 1
Foundations: How LLM Coding Agents Work
4 lessons
- 01From Chatbots to Agents: What Changes in the Terminal12 min
- 02The Anatomy of a Request: System Prompt, User Prompt, and Tool Calls13 min
- 03Models Under the Hood: Tokens, Temperature, and Reasoning13 min
- 04Setting Up Your First Agent Workspace12 min
Module 2
Prompting and Context Management
4 lessons
- 01Writing Effective System Prompts and Project Instructions13 min
- 02Managing the Context Window: What Fits and What Gets Cut13 min
- 03Project Memory: Docs, Indexes, and Retrieval14 min
- 04Prompting Patterns That Get Better Code12 min
Module 3
Agent Workflows: Planning, Multi-Step Tasks, and Tool Use
4 lessons
- 01Planning Mode: Letting the Agent Think Before It Acts13 min
- 02Tool Use in Practice: Files, Shell, and Search13 min
- 03Orchestrating Multi-Step Tasks and Long-Running Sessions14 min
- 04Common Agent Failure Modes and How to Recover14 min
Module 4
Code Review, Refactoring, and Debugging with AI
3 lessons
- 01AI-Assisted Code Review: What to Ask and What to Trust13 min
- 02Refactoring at Scale with Agentic Tools13 min
- 03Debugging with Agents: From Stack Traces to Root Cause14 min
Module 5
Security, Guardrails, and Integration
4 lessons
- 01Permissions, Sandboxing, and Safe Tool Use14 min
- 02Secrets, Supply Chain, and Prompt-Injection Risks15 min
- 03Integrating Agents into Git and CI/CD Workflows14 min
- 04Choosing Your Stack: Claude Code vs. Codex vs. GPT Tools14 min