How-To Guides

Codex CLI Commands: The Essential Reference for PowerShell and WSL

The Codex CLI commands worth knowing, on one page: permission flags including --yolo, resume, and the slash commands for planning, context and review. Written for PowerShell and WSL users.

  • ️ Permission flags explained, including --yolo and --sandbox danger-full-access
  • Resume a session with codex resume instead of starting over
  • codex doctor and tab completion for troubleshooting
  • ️ The slash commands that matter for planning, context and review
By Rejith Krishnan3 min read
Codex CLI Commands: The Essential Reference for PowerShell and WSL

Installing Codex takes an hour. Learning what it can do takes longer, because most of the useful commands never show up on screen. This guide lists the ones worth knowing.

If Codex is not installed yet, start with a setup guide and come back:

Commands and flags below were checked against codex --help. Codex ships often, so run codex --version and codex <command> --help if something does not match your install.

Starting Codex

codex

Permissions

codex --dangerously-bypass-approvals-and-sandbox
codex --yolo
codex --sandbox danger-full-access

--yolo is the short alias of --dangerously-bypass-approvals-and-sandbox. Both skip every approval prompt and every sandbox, so use them only inside a container or VM that is already isolated. --sandbox danger-full-access removes the sandbox but keeps the approval policy.

Inside a session, /permissions changes the same settings without a restart.

For more flag details, check codex --help.

Resume a session

Codex saves every interactive session.

codex resume            # pick a previous session from a list
codex resume --last     # reopen the most recent session

Troubleshooting

  • codex doctor checks the install, config, auth and network. Run it first when Codex will not start or connect. codex doctor --json prints a redacted report you can attach to a ticket.
  • codex completion powershell (Windows) or codex completion bash (WSL) generates tab completion.

Slash commands inside a session

Type / in an empty composer to open the menu, and keep typing to filter it.

Model

Command What it does
/model Choose the model and reasoning effort
/fast Run at 1.5x speed, with higher usage

Planning and context

Command What it does
/init Create an AGENTS.md file with instructions for Codex
/plan Switch to Plan mode
/goal Set or view the goal for a long-running task
/compact Summarize the conversation to stay under the context limit
/side (alias /btw) Start a side conversation in a throwaway fork
/new Start a new chat

Code and review

Command What it does
/review Review your current changes and find issues
/diff Show the Git diff, including untracked files

Tools and integrations

Command What it does
/mcp List MCP tools; /mcp login <name> to sign in
/plugins Browse plugins
/skills Use skills to improve how Codex handles specific tasks
/hooks View and manage lifecycle hooks

What's next?

These are the commands our engineers use when they run Codex inside Krithi, the lowtouch.ai SDLC agent that takes a ticket through planning, coding, testing and a compliance audit to a reviewed pull request, with a named human approving each stage.

About the Author

Rejith Krishnan

Rejith Krishnan

Founder and CEO

Rejith Krishnan is the Founder and CEO of lowtouch.ai, a platform dedicated to empowering enterprises with private, no-code AI agents. With expertise in Site Reliability Engineering (SRE), Kubernetes, and AI systems architecture, he is passionate about simplifying the adoption of AI-driven automation to transform business operations.

Rejith specializes in deploying Large Language Models (LLMs) and building intelligent agents that automate workflows, enhance customer experiences, and optimize IT processes, all while ensuring data privacy and security. His mission is to help businesses unlock the full potential of enterprise AI with seamless, scalable, and secure solutions that fit their unique needs.

LinkedIn →