NEW Codex CLI Update Is ABSURD!
OpenAI's new Codex CLI update transforms terminal-based programming assistance by enabling multiple AI agents to work simultaneously on different tasks, visible in a full-screen control center interface. The update includes features like Agent View, work trees, voice control, and parallel task execution, fundamentally changing how AI assistance operates from sequential Q&A to autonomous team-based work management.
Summary
Julian Goldie presents a detailed analysis of OpenAI's latest Codex CLI update, which represents a significant shift in how AI tools operate. Previously, AI tools functioned in a simple linear fashion: users typed questions and received answers sequentially. The new update enables multiple AI agents to work in parallel on different parts of a larger task, which users can monitor simultaneously on a single screen.
The update introduces several key features. Full-screen mode replaces the previous narrow text strip at the bottom of the screen, now displaying the entire chat with AI, tools being used, files changed, and task status in one unified interface. Agent View functions as a control room, allowing users to monitor multiple agents and their work simultaneously—similar to a supervisor watching multiple kitchen chefs execute different tasks rather than one chef working sequentially.
The update includes tools for multi-agent coordination, allowing agents to open, read, or fork (branch) to other tasks. This enables a structured team approach where a lead agent, researcher, developer, tester, and reviewer can each handle different aspects of a project. Work trees provide separate workspaces for each task using Git version control, preventing conflicts when multiple agents modify files simultaneously. Additional features include voice control for communication, visual displays like charts directly in the terminal, and usage statistics.
Goldie emphasizes this represents a genuine operational shift in how work gets done, evidenced by version progression (0.156 to 0.157.0) and OpenAI developer documentation. He stresses the importance of verification—that while this is significant, AI still makes mistakes and requires human oversight. The update's value lies in visibility and control, allowing users to see what AI is doing in real-time.
Key Insights
- OpenAI transformed Codex from a simple sequential question-answer tool into a platform where multiple AI agents can work simultaneously on different parts of a task while users watch all execution on one screen, fundamentally changing the workflow from linear to parallel task management.
- Agent View functions as a control room metaphor—similar to one supervisor monitoring multiple kitchen chefs each with assigned tasks rather than a single chef working sequentially—enabling visual oversight of multiple agents working in parallel.
- Work trees solve the problem of multiple agents modifying the same files simultaneously by giving each agent their own separate workspace using Git version control, then consolidating results afterward to prevent conflicts and chaos.
- According to the Codex GitHub page, full-screen mode was iteratively developed from version 0.156 to 0.157.0, where it became standard, demonstrating this update emerged through deliberate step-by-step engineering rather than sudden innovation.
- Goldie argues there is a limited window of opportunity for business owners to gain competitive advantage with AI agents, as most people currently use AI for simple sequential tasks and few understand multi-agent team management, but this gap will close as adoption spreads.
Topics
Transcript
[0:00] The new Codex CLI update is simply absurd. And the reason is simple. OpenAI has just transformed the Codex terminal into a place where you can watch multiple AI agents work simultaneously. Let me say this straight. You are giving Codex a big task. He breaks this task down. Different agents take on different parts. And you sit and watch it all happen on one screen. A year ago, most AI tools worked like this. You typed a question, got an answer , and then typed the next one. Now you can assign an entire job to AI and just occasionally check the status of its execution. This is a completely different way of working. [0:30] OpenAI's help pages…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Julian Goldie SEO
Google AI Studio Just Got A Huge Security Upgrade
Google is reportedly developing a new "Security Check" mode for AI Studio that would apply secure coding practices to AI-generated applications. The feature aims to automate security verification during app development, addressing common vulnerabilities like exposed keys, weak authentication, and data access issues before deployment.
New Gemini Skills Update Is WILD!
Google has launched Gemini Skills, a new feature allowing users to save custom instructions once and reuse them repeatedly by typing a slash command, replacing the discontinued Gems feature. Skills can include reference files, be combined together, and are built on an open file format (skill.md) that allows portability across different AI tools.
NEW Hermes Cloud + Agent OS is Crazy! 🤯
Hermes Cloud combined with Agent OS enables users to deploy AI agents in the cloud 24/7 using free models without API keys or local setup complexity. The system provides persistent memory, sandbox isolation, and integration with custom workflows like content research, competitor analysis, and news monitoring.
Manus 2.0 Just Changed AI Agents Forever
Manas 2.0 represents a fundamental architectural overhaul that enables users to create games, websites, videos, and automations through simple text requests, with the key innovation being editable components rather than fixed outputs. The platform uses a new Cascade engine that reduces token usage by 23% and improves task speed by 28%, while maintaining all project elements in one connected workspace.
How to Run DeepSeek V4 Flash for FREE!
A tutorial demonstrating how to use DeepSeek V4 Flash for free through Open Code and integration with agent operating systems like Hermes Agent. The speaker showcases building websites and apps using this free AI model and explains how it compares favorably to larger models despite being smaller.