OpenAI Expands Codex With Reusable Cloud Workspaces Across Devices

OpenAI is expanding Codex with reusable cloud workspaces, voice controls and repository security tools to let developers manage software tasks across computers, phones and the cloud.

The company announced the capabilities Sept. 29 as part of a broader Codex update. Reusable development environments can launch tasks faster and provide shared workspaces with approved settings and permissions. Unlike isolated cloud tasks, they can persist and be configured for repeated use.

The Codex command-line interface is also gaining voice controls and a new /agents view for delegating and tracking multiple tasks. Developers can edit prompts, resume sessions and use worktrees through the updated interface, which also includes a redesigned terminal experience for longer sessions.

Codex is being added to a code review experience in the ChatGPT desktop app. Users can review summaries, examine code changes and ask about potential issues before sharing feedback on GitHub pull requests or GitLab merge requests. Automatic reviews can run while users are away from their computers.

OpenAI is also introducing Codex Security Cloud, a set of repository scanning tools that can inspect entire GitHub repositories on demand, on a schedule or after new commits. The system investigates findings, removes duplicate results and prepares fixes in the cloud, including while the user is inactive and the laptop is closed.

The security tools also include access to models offered through OpenAI’s Daybreak Blue cybersecurity initiative without requiring a separate application.

Additional API updates include a Decisions API that uses Luna to answer user-defined questions with preset responses. An updated Agents API supports computer use and Amazon Bedrock Managed Agents, allowing OpenAI agents to run entirely on AWS.