Course : Developing with OpenAI Codex

Master the OpenAI Codex development tool to automate code generation, code reviews, and DevOps workflows.

Practical course - 2d - 14h00 - Ref. DOI
Price : 1500 CHF E.T.

Developing with OpenAI Codex

Master the OpenAI Codex development tool to automate code generation, code reviews, and DevOps workflows.


New course

This training course is designed to help developers understand and use OpenAI Codex as an advanced programming assistant. Participants will learn how to generate, analyze, and refactor code through conversational interactions; automate development tasks; integrate Codex into their work environments and CI/CD pipelines; and design AI-enhanced development workflows to improve productivity and software quality.


INTER
IN-HOUSE
CUSTOM

Practical course in person or remote class
Disponible en anglais, à la demande

Ref. DOI
  2d - 14h00
1500 CHF E.T.




This training course is designed to help developers understand and use OpenAI Codex as an advanced programming assistant. Participants will learn how to generate, analyze, and refactor code through conversational interactions; automate development tasks; integrate Codex into their work environments and CI/CD pipelines; and design AI-enhanced development workflows to improve productivity and software quality.


Teaching objectives
At the end of the training, the participant will be able to:
Install and configure Codex CLI and the VS Code IDE extension
Effectively use the various interaction modes (interactive, exec, cloud)
Create optimized AGENTS.md files to guide the agent
Configure MCP servers to expand Codex's capabilities
Integrating Codex into CI/CD pipelines with GitHub Actions
Automate code reviews and repetitive tasks
Implement best practices for security and sandboxing

Intended audience
Developers (all languages), DevOps and SRE engineers, software architects, tech leads, and technical team leads.

Prerequisites
Experience in software development (2+ years recommended), knowledge of Git and GitHub/GitLab workflows, familiarity with the terminal/command line. An active ChatGPT Plus, Pro, Business, or Enterprise account.

Practical details
A mix of theoretical presentations and coding workshops. Each concept is immediately applied to real-world scenarios.
Teaching methods
The instructor assesses the participant’s learning progress throughout the training using multiple-choice questions, role-playing exercises, hands-on activities, and more. The participant also completes a placement test before and after the training to validate the skills acquired.

Course schedule

1
Introduction to the Codex Ecosystem

  • Codex's Position in the OpenAI Ecosystem.
  • Architecture: CLI, IDE extension, cloud, integrations.
Hands-on work
Overview of Codex interfaces. First interaction with the agent. Installation via npm or Homebrew.

2
An In-Depth Look at Codex CLI

  • Essential keyboard shortcuts.
  • Navigating the terminal interface.
Hands-on work
Explore an unfamiliar project with Codex. Ask for explanations about the architecture. Generate documentation automatically. Work on a task across multiple sessions. Use the previous context to continue.

3
Control and Safety

  • Manage approval modes.
  • The three levels: Read-only, Auto (default), Full Access.
Hands-on work
Test the different permission levels. Observe how the agent behaves depending on the mode. Test the sandbox restrictions. Configure targeted access.

4
AGENTS.md and Customization

  • Purpose of the AGENTS.md file: a README for agents.
  • Create an AGENTS.md file for the Fil Rouge project.
Hands-on work
Write an effective AGENTS.md file for building a simple application.

5
IDE and Cloud Extension

  • The VS Code extension.
  • Modes: Chat, Agent, Agent (Full Access).
Hands-on work
Install and configure the extension. Codex-assisted refactoring. Generate unit tests. Code navigation and explanation.

6
MCP and GitHub Integration

  • Understanding the MCP: Connecting Models to External Tools.
  • Configuration in config.toml.
Hands-on work
Set up an MCP documentation server (Context7). Add a custom MCP server using `codex mcp add`. Use MCP tools in a Codex session. GitHub: Enable automatic review on a PR. Customize review guidelines.



This programme is an original creation, developed by the teaching teams at ORSYS Formation. Any reproduction, representation, adaptation or use, in whole or in part, without the prior written authorisation of ORSYS, is strictly prohibited. ORSYS reserves the right to take any action necessary to protect its intellectual property rights.

Dates and locations

Dernières places
Date garantie en présentiel ou à distance
Session garantie
From 17 to 18 December 2026
FR
Remote class
Registration
From 1 to 2 March 2027
FR
Remote class
Registration
From 13 to 14 May 2027
FR
Remote class
Registration
From 7 to 8 June 2027
FR
Remote class
Registration
From 12 to 13 July 2027
FR
Remote class
Registration
From 22 to 23 November 2027
FR
Remote class
Registration
From 13 to 14 December 2027
FR
Remote class
Registration

REMOTE CLASS
2026 : 17 Dec.

2027 : 1 Mar., 13 May, 7 June, 12 July, 22 Nov., 13 Dec.



This programme is an original creation, developed by the teaching teams at ORSYS Formation. Any reproduction, representation, adaptation or use, in whole or in part, without the prior written authorisation of ORSYS, is strictly prohibited. ORSYS reserves the right to take any action necessary to protect its intellectual property rights.