OpenAI makes Codex cloud environments reusable across devices
OpenAI has expanded its Codex software engineering agent with reusable cloud development environments that can be accessed from different devices. The company announced the changes at Dev Day, alongside a refreshed Codex CLI, a new code review experience, infrastructure security tools and API updates.
Unlike earlier cloud tasks that operated as separate remote sandboxes, the new environments are designed to persist and be configured for repeated use. OpenAI said they can help tasks begin faster and provide teams with shared workspaces containing approved settings and permissions. The updated CLI also supports starting and directing tasks by voice, while a new /agents view is intended to help users delegate and monitor several tasks. Other changes include session resumption, prompt editing, worktree support and a revised terminal interface for longer sessions.
Within the ChatGPT desktop app, Codex can summarize code changes, investigate possible issues and support reviews before feedback is posted to GitHub pull requests or GitLab merge requests. Codex Security Cloud can scan GitHub repositories on demand, on a schedule or after new commits, then investigate findings, remove duplicates and prepare fixes in the cloud. OpenAI also introduced the Decisions API, which uses Luna for user-defined questions with predefined answers, and updated Agents API support for computer use and Amazon’s Bedrock Managed Agents.
Written by Sakhanda Wire AI from reporting by TechCrunch. Facts, quotes and full context are in the original article below.