# Shipping with Codex Source: https://www.youtube.com/watch?v=Gr41tYOzE20 Recap page: https://rapidrecap.app/video/Gr41tYOzE20 Generated: 2025-10-08T17:36:01.979+00:00 --- ## Quick Overview Codex is evolving into a proactive, powerful AI software engineer teammate, demonstrated by its 10x usage increase from August to October and its integration into internal workflows like code reviews and planning, allowing engineers to ship code faster and with more confidence by handling complex tasks and generating necessary documentation and tests. **Key Points:** - Codex usage saw a massive 10x shift/increase between August and October due to its integration into daily engineering workflows. - Codex is viewed as a collaborative partner, async agent, and proactive teammate, capable of tasks like pairing on code and delegating work without explicit prompting. - The Codex Agent architecture consists of a 'Harness' and a 'Model' (like GPT-5 Codex), allowing it to work across various environments (IDE, Terminal, GitHub, Cloud, Mobile). - OpenAI engineers are now using Codex to generate test suites (like unit tests) and documentation (like design docs) for complex features, significantly accelerating development. - Internal metrics show 92% of OpenAI technical staff use Codex daily, 70% more PRs are submitted per week by Codex users, and 100% of PRs are reviewed by Codex. - The development process emphasizes iterative feedback, where Codex helps generate code, tests, and documentation, which are then reviewed, leading to faster, higher-quality shipping. ![Screenshot at 00:07: Tibo Sottiaux introduces the session on "Shipping with Codex," highlighting its role in development acceleration.](https://ss.rapidrecap.app/screens/Gr41tYOzE20/00-00-07.png) **Context:** This presentation, titled "Shipping with Codex," features OpenAI employees Tibo Sottiaux and Aaron Friel discussing the evolution, adoption, and practical application of Codex within OpenAI's own engineering processes. The focus is on how Codex has moved beyond simple code completion to become an essential, proactive AI teammate that accelerates development cycles, particularly in areas like code review and feature planning. ## Detailed Analysis The presentation details the rapid evolution of Codex, positioning it as an AI software engineer teammate rather than just a junior engineer. Tibo Sottiaux highlighted a massive 10x usage increase between August and October, driven by its integration across IDEs, terminals, GitHub, and the cloud. He explained the Codex Agent architecture, which separates the 'Harness' (tools/workflow) from the 'Model' (like GPT-5 Codex). This agent structure allows Codex to perform complex tasks like generating code, test suites, and documentation based on context, as demonstrated by generating a comprehensive design document for a new weather app feature. Aaron Friel followed up by showcasing how Codex is used internally for code reviews, noting that 92% of technical staff use it daily, leading to 70% more PRs submitted weekly and 100% of PRs receiving review suggestions from Codex. He demonstrated using Codex to plan, implement, and ship a feature, including generating tests and documentation, emphasizing that this process is now faster and results in higher-quality code due to the AI's ability to handle complex tasks and provide iterative feedback, effectively acting as a senior engineer. ### Codex Adoption & Impact - Usage grew 10x between August and October - 92% of OpenAI staff use Codex daily - 100% of PRs reviewed by Codex ### Codex Agent Architecture - Composed of a 'Harness' (tools/workflow) and a 'Model' (like GPT-5 Codex) - Supports IDE, Terminal, GitHub, Cloud, and Mobile environments ### Internal Use Case - Accelerating development by using Codex for complex feature implementation, including generating test suites and design documents ### Code Review Workflow - Codex automatically suggests improvements on PRs; users can select presets like 'Review against a base branch' or 'Review a commit' ### Planning & Iteration - Engineers use an ExecPlan (living document) to structure work, which Codex can help populate and verify, ensuring rigorous planning and testing. ![Screenshot at 00:00: Title card showing the event: OpenAI DevDay \[2025\].](https://ss.rapidrecap.app/screens/Gr41tYOzE20/00-00-00.png) ![Screenshot at 00:07: Tibo Sottiaux begins his presentation on "Shipping with Codex," identifying himself as being from OpenAI.](https://ss.rapidrecap.app/screens/Gr41tYOzE20/00-00-07.png) ![Screenshot at 00:15: Slide illustrating Codex's role as a 'Collaborative partner,' 'Async agent,' and 'Proactive teammate.'](https://ss.rapidrecap.app/screens/Gr41tYOzE20/00-00-15.png) ![Screenshot at 00:37: Slide showing a 10x increase in usage from August to October, highlighting massive adoption.](https://ss.rapidrecap.app/screens/Gr41tYOzE20/00-00-37.png) ![Screenshot at 00:56: Slide outlining the presentation structure: 'Rapid evolution of Codex' and 'Codex inside OpenAI engineering.'](https://ss.rapidrecap.app/screens/Gr41tYOzE20/00-00-56.png) ![Screenshot at 01:25: Slide detailing the Codex Agent's capability across various tools: IDE, Terminal, GitHub, Cloud, and Mobile.](https://ss.rapidrecap.app/screens/Gr41tYOzE20/00-01-25.png) ![Screenshot at 01:51: Diagram showing the Codex Agent composition: 'Harness' and 'Model' \(GPT-5 Codex\).](https://ss.rapidrecap.app/screens/Gr41tYOzE20/00-01-51.png) ![Screenshot at 02:37: A slide emphasizing that Codex is not a 'Junior Engineer' and that previous attempts at similar tools failed.](https://ss.rapidrecap.app/screens/Gr41tYOzE20/00-02-37.png) ![Screenshot at 03:33: A slide showing a complex set of changes in a Git diff, illustrating the scale of work Codex can handle.](https://ss.rapidrecap.app/screens/Gr41tYOzE20/00-03-33.png) ![Screenshot at 04:25: A screen showing 'Codex cloud tasks' on a mobile interface, indicating remote capability/management of tasks against the code base shown in the previous screen's diff view \(implied\). \(Note: The screen is dark, but the context suggests this is the cloud task view\). A faint screenshot of a mobile task list is visible on the right of the main screen area in this shot, though the main screen is dark in this frame. The visual context is primarily the speaker and the dark screen with text overlay 'Codex cloud tasks'. The visual element on the right of the screen is a mobile UI snippet showing a list of tasks, confirming the cloud task capability mentioned by the speaker in the surrounding audio \(04:28\). The visual evidence in the screenshot is faint but correlates with the audio mention of cloud tasks and mobile interaction \(04:31\). I will rely on the audio context for the specific detail, but the visual confirms the environment change.](https://ss.rapidrecap.app/screens/Gr41tYOzE20/00-04-25.png)