News
eBook
Designing an AI-Native Contact Center
OpenAI today introduced the Codex app for macOS, a desktop application designed to help developers manage multiple AI agents at once, run parallel workstreams and oversee complex software projects from start to finish.
Rather than pairing with a single assistant for quick edits, teams can now supervise coordinated agents working across the full lifecycle of software design, development, deployment and maintenance.
For a limited time, Codex will be included with ChatGPT Free and Go plans, while rate limits for Plus, Pro, Business, Enterprise and Edu subscribers doubled across the app, CLI, IDE extensions and cloud usage.
Table of Contents
- Codex Is the New AI Agent Command Center
- Moving Beyond Code Generation With 'Skills'
- Built-In Skill Library Integrations
- Automations to Offload Repetitive Work
- Is the Codex App Secure?
- Growing Adoption and What’s Next
Codex Is the New AI Agent Command Center
Since Codex first launched in April 2025, OpenAI says developer workflows have evolved significantly. AI agents are increasingly capable of handling complex, long-running assignments, sometimes spanning hours, days or even weeks.
But now there's a new bottleneck: how to direct, supervise and collaborate with multiple agents at scale.
The Codex desktop app is designed to address that challenge by serving as a centralized workspace where developers can:
- Run multiple agents in parallel
- Organize tasks by project
- Switch between threads without losing context
- Review and comment on code changes
- Open agent-generated diffs directly in their editor
Each agent operates in its own thread, with built-in support for worktrees. That allows multiple agents to work on the same repository simultaneously without creating merge conflicts.
Key capabilities include:
- Isolated code copies for each agent
- Local checkout of agent changes
- Parallel experimentation across different solution paths
- Seamless continuation of work without disrupting local Git state
The app also syncs session history and configurations from the Codex CLI and IDE extensions, allowing developers to resume work on existing projects.
Related Article: OpenAI, The ‘Bailout,' and the Likely Path Forward
Moving Beyond Code Generation With 'Skills'
Codex now supports “skills” — modular bundles of instructions, scripts and resources that allow Codex to perform broader tasks beyond code generation.
Skills enable Codex to:
- Connect to third-party tools
- Execute predefined workflows
- Synthesize information
- Solve problems across knowledge work domains
Within the Codex app, developers can create, manage and share skills across teams. Skills can be invoked explicitly or automatically triggered based on task requirements.
Example: Building a Racing Game End-to-End
OpenAI demonstrated Codex’s capabilities by asking it to build a racing game featuring:
- Multiple racers
- Eight maps
- Interactive game mechanics
Using an image generation skill powered by GPT Image and a web game development skill, Codex completed the project independently. According to OpenAI, the agent processed more than 7 million tokens from a single initial prompt, assuming roles such as designer, developer and QA tester (which included playing the game to validate functionality).
Built-In Skill Library Integrations
The Codex app includes a library of pre-built skills for popular workflows and tools. Examples include:
| Workflow Area | Integrated Tools & Capabilities |
|---|---|
| Design Implementation | Pull assets and context from Figma and generate production-ready UI code |
| Project Management | Triage and manage releases in Linear |
| Cloud Deployment | Deploy apps to Cloudflare, Netlify, Render and Vercel |
| Image Generation | Generate and edit visuals using GPT Image |
| OpenAI API Development | Reference up-to-date OpenAI API documentation |
| Document Creation | Create and edit PDF, spreadsheet and DOCX files |
Teams can also check skills into repositories, making them reusable across organizations.
Automations to Offload Repetitive Work
Beyond real-time collaboration, the Codex app introduces Automations, scheduled background tasks that combine instructions with optional skills.
Developers can configure Codex to:
- Run on defined schedules
- Execute recurring workflows
- Deliver results into a review queue
OpenAI says internal teams use Automations for:
- Daily issue triage
- CI failure summaries
- Release brief generation
- Bug checks
Future updates will expand Automations to support cloud-based triggers, allowing Codex to run continuously in the background.
Is the Codex App Secure?
OpenAI says the Codex app integrates system-level sandboxing similar to its CLI implementation. By default, agents are restricted to:
- Editing files within their designated folder or branch
- Using cached web search
- Requesting permission before executing elevated commands such as network access
Teams can configure project-level rules that allow specific commands to run automatically with elevated permissions.
Related Article: Vibe Coding: Reimagining Software Development for the Age of Agents
Growing Adoption and What’s Next
OpenAI reports that since the mid-December launch of GPT-5.2-Codex, overall Codex usage has doubled. More than one million developers have used Codex in the past month.
Webinar
Feb
25
Content Strategy Leaders Live: Scaling for Speed, Complexity and AI in High Tech
A candid roundtable on how high-tech leaders are rethinking content at scale.
Register
Webinar
Feb
26
Do More with Less: Modernizing the Cloud Contact Center for 2026
Learn how to leverage cloud platforms without adding a single hire to personalize every customer interaction.
Register
Webinar
On demand
Cut the Noise: Deploying AI That Actually Moves the Needle
Learn how to turn AI experimentation into concrete revenue operations.
Watch Now
Webinar
On demand
Ditch the Desk Phones: How Modern Teams Drive AI-First Communications
Find out how one team finally pulled the plug on a legacy phone system. And built something smarter.
Watch Now
Webinar
On demand
Rebrand. Migrate. Optimize. How to Do It All (Without Slowing Down)
Cresta leveled up site speed, design flexibility and marketer sanity (in record time). Find out how.
Watch Now
Webinar
On demand
From Manual to Magical: How AI Transforms CX Teams
Learn how to replace manual support processes with automation that actually delivers.
Watch Now
Looking ahead, OpenAI plans to:
- Release a Windows version of the app
- Improve multi-agent workflow management
- Expand automation capabilities
- Deliver faster inference performance
The broader goal, according to OpenAI, is to close the gap between what advanced AI models are capable of and how easily developers can apply them to real-world tasks.