Elena' s AI Blog

Codex CLI Part 2 — Security Controls & Safe Editing

06 Feb 2026 (updated: 24 Aug 2026) / 35 minutes to read

Elena Daehnhardt


Cloud and AI systems illustration


TL;DR:
  • Codex uses permissions (what it can do) and approvals (when it asks). Three modes combine both: Read Only (safe exploration), Default/Agent (balanced), Full Access (trusted tasks only). Use /approvals or /permissions to switch modes. Always /diff before accepting changes. Part 2 of 4.

Previous: Part 1 — Getting Started with Codex CLI

Next: Part 3 — Codex CLI Part 3: Practical Workflows for Blogging and Python Development

This is Part 2 of the Codex CLI series. Today, we’ll learn how to control Codex’s capabilities and make your first safe edits.

This post is about speed with control, not automation for its own sake.

Codex CLI Part 2: Controlling Permissions and Making Safe Edits

Codex CLI is a terminal-based coding agent from [OpenAI][1] that can read, edit, and run commands directly in your local development environment, rather than only suggesting changes like a chatbot. In the last post, we installed Codex CLI and took our first steps with read-only exploration. Today, we go deeper: learning how to control what Codex CLI can do, understanding the essential commands you will use daily, and making your first actual code edits—safely.

Codex CLI’s productive use depends on understanding its control mechanisms. Unlike a chatbot that only gives suggestions, Codex CLI can actually change your files and run commands. That power is valuable, but it requires proper guardrails.

Let me show you how to stay in control while getting real work done.

Codex CLI Security Model: Permissions vs. Approvals

Permissions are the technical boundaries that define what Codex CLI is capable of doing, and approvals are the workflow checkpoints that determine when Codex CLI must pause and ask for your explicit consent. Understanding the distinction between the two is the most important concept before you let Codex CLI make any changes.

At first, permissions and approvals might seem like the same thing, but understanding the distinction helps you think clearly about how Codex CLI operates and how to stay in control.

Permissions vs Approvals: The Conceptual Difference

Permissions: WHAT Codex Can Do

Permissions define the technical boundaries—what Codex is actually capable of doing in your environment:

  • Network Access: Can Codex connect to the internet, external APIs, or package registries?
  • File Access Scope: Can Codex only read? Edit workspace files? Edit files anywhere on your system?
  • Command Execution: Can Codex run shell commands, install packages, or execute scripts?

Think of permissions as the capabilities Codex has been granted. They are the “can it” question.

Approvals: WHEN Codex Must Ask

Approvals determine the workflow control—when Codex needs to pause and get your explicit consent before taking action:

  • Should Codex ask before accessing the network?
  • Should Codex ask before editing any file?
  • Should Codex proceed autonomously within its granted permissions?

Think of approvals as the checkpoints where you maintain control. They are the “must it ask first” question.

Essential Slash Commands

🔒 Subscribe to keep reading.

Making Your First Safe Edit

🔒 Subscribe to keep reading.

Codex CLI Best Practices for Safe, Auditable Edits

🔒 Subscribe to keep reading.

Codex CLI Practice Exercise: The Safe Edit Challenge

🔒 Subscribe to keep reading.

Codex CLI Series Roadmap: What’s Covered in Part 3

🔒 Subscribe to keep reading.

Codex CLI Security Model Summary: Key Takeaways

🔒 Subscribe to keep reading.

Subscribe to unlock the full article ❤️

I keep most of the site completely open. A few unusually detailed tutorials need a free subscriber login so I can keep publishing this kind of work.

The form below signs you up for the newsletter. It does not log you into the app — log in afterwards (same email) to unlock this article and download your subscriber gifts. New subscribers get an inbox mail: Set a password to unlock articles.

desktop bg dark

About Elena

Elena, a PhD in Computer Science, simplifies AI concepts and helps you use machine learning.




Citation
Elena Daehnhardt. (2026) 'Codex CLI Part 2 — Security Controls & Safe Editing', daehnhardt.com, 06 February 2026. Available at: https://daehnhardt.com/blog/2026/02/06/codex-cli-part-2-security-controls-and-safe-edits/
All Posts