OpenAI Codex: Your First Steps

OpenAI's guide to its Codex AI coding assistant details a new desktop app, project-based workflow, and recommended setup for developers.

Screenshot of the OpenAI Codex desktop application interface showing a chat thread and project sidebar.
The OpenAI Codex desktop app interface for AI-assisted coding.· OpenAI News

OpenAI is rolling out new tools to make its AI coding assistant, Codex, more accessible. The company's latest OpenAI News post details a streamlined approach for developers to get started.

The process begins with downloading the Codex desktop app and logging in via your ChatGPT account. Upon opening, users create a 'thread,' which functions as a conversational workspace for task completion.

Codex organizes projects and conversation history in a sidebar, with the main area dedicated to the current thread. Projects are directly linked to local folders on your computer, allowing Codex to interact with specific files or create new ones within designated directories.

Setting Up Your First Project

To manage projects efficiently, OpenAI suggests creating a dedicated 'Codex' folder on your machine. Within this, each project can have its own subfolder. Simply drag relevant project files into the folder to make them accessible to Codex.

For initial tasks, start with something straightforward and practical, such as organizing notes, cleaning a small dataset, or comparing document drafts.

When working within a project, you'll encounter the 'Work locally' option, indicating Codex's operational scope is confined to the designated folder and chosen tools. Permissions can be adjusted, but the recommendation for beginners is to stick with 'Default permissions' in a local environment.

Codex does not have unfettered access to your entire system by default. 'Full permissions' offer greater operational freedom for advanced tasks but require careful consideration and administrative approval.

A recommended first prompt is: "Inspect this folder and tell me what you see. Then suggest one small task you can complete safely. Wait for my approval before making changes."

This iterative approach of starting small, reviewing output, and gradually increasing task complexity is key to building confidence and trust in the AI's capabilities.

For those looking to deepen their understanding, OpenAI Academy offers additional guides and resources on practical AI skill development.

© 2026 StartupHub.ai. All rights reserved. You may not republish this article in full without a license. Search engines and AI research tools may crawl and summarize for reference. Bulk reproduction or model training requires a license. See our terms.
Daniel Singer

Written by

Daniel Singer

Editor, StartupHub.ai

Daniel Singer is the editor of StartupHub.ai, a technology expert and thought leader on AI and its applications across sectors, from fintech and healthcare to developer tooling and consumer software. He writes and tests the tools covered here thoroughly and regularly, and built StartupHub.ai to give founders, operators and buyers a clearer read on what they are actually being sold.