1
Sandbox creation
Codex provisions a fresh, isolated environment for your task. Each task gets its own clean environment.
2
Repository clone
Your GitHub repository is cloned into the sandbox. Codex gets a full copy of your codebase to work with.
3
Environment setup
Dependencies are installed and tools are configured based on your AGENTS.md file or detected project configuration.
4
Task execution
Codex reads code, plans changes, writes new code, and runs your test suite to verify everything works.
5
Pull request creation
Changes are committed, pushed to a new branch, and a pull request is opened on GitHub for your review.
0M+
Weekly users
0
Product surfaces
0%
Isolated execution
0
PR per completed task
Cloud vs. local execution: what is the difference?
Local tools like Claude Code and the Codex CLI run directly on your computer. They can access your files, databases, local servers, and terminal. Cloud tasks run on OpenAI infrastructure in isolated sandboxes. This is safer (nothing on your machine is touched) but also means no access to local resources. The Codex desktop app and CLI give you local execution when you need it, while Codex Cloud handles background tasks that benefit from isolation. Many developers use both: local for interactive work, cloud for background tasks.
Triggering tasks from GitHub
Tag @codex on any GitHub issue or pull request to spin up a cloud task. Codex reads the context, works on the task in an isolated sandbox, and delivers the results as a pull request. This lets you assign work to Codex without leaving GitHub. Cloud tasks also integrate with Slack and Linear for notifications.