The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
Codex CLI's edge: open-source harness + OS sandbox
I ship client code with Claude Code. When I evaluate Codex as a second seat, the interesting part isn't another model race — it's that the harness itself is open-source and the OS sandbox is explicit.
OpenAI publishes Codex as a lightweight terminal coding agent (github.com/openai/codex). Sandbox modes are separate from approval prompts: read-only inspects files and asks before edits or commands; workspace-write (the default low-friction path) reads and edits inside the workspace with routine local commands inside the boundary and network restricted unless you enable it; danger-full-access lifts the sandbox. Full access = danger-full-access + approvals set to never. Docs: https://learn.chatgpt.com/docs/sandboxing.md
Project instructions live in AGENTS.md — global ~/.codex/AGENTS.md plus nested project files — so a second seat can inherit the same house rules without a private config format.
How I'd try it as seat 2: start workspace-write + on-request, keep Claude Code on the hard multi-file ship path, and only open danger-full-access for a throwaway sandbox repo.
Do you want your coding CLI open-source even if you still pay for the model behind it?
More at huzaifatahir.dev
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Related posts
Did a motion for the Cockpit concept ⁕ 🔈 ON
Design & Creative Direction: Me
Motion & Sound: Claude Code
Cockpit is an AI agent management platform concept that helps teams:
- Set boundaries
- Monitor activity
- Manage approvals
- Track knowledge sources
- Measure performance and costs
It brings together five key AI UX challenges:
👉 Autonomy
👉 Observability
👉 Human Approvals
👉 Context
👉 Analytics
The idea is simple:
As AI agents become more capable, the experiences we design around them need to make their actions more transparent, predictable, and manageable.
Thanks!
Human approvals is the part most agent UIs skip. Good call making it one of the five challenges.
Weekend fun 😺. The new MacBook Pro has a Dynamic Island, and I kept wondering: what if that little pill was where your AI agent lived?
Try it live: https://lab.behzad.me/agent-mew/
So I spent the weekend prototyping Agent Mew. You drop a file on the notch, give it a task, and it quietly works up there while you keep doing your thing. It shows progress in the wings, asks a quick question when it needs one, and always waits for your OK before doing anything that leaves your Mac, like sending an email.
A few things I explored:
- Progress that stays out of the way, readable at a glance
- Questions as one tap, not a chat window
- Approval cards for anything with real consequences
- Paused and stuck states that tell you what to do next
- A small animated cat, because agents should have some personality
Approval cards for anything that leaves the Mac is the right line to draw. Does the card show the full email before you tap OK, or a short summary? I'd lean toward full text for anything going to another person.
Designed an AI research workspace where the agent plans the work and fills a sheet with cited sources as it goes.
Same screen, two modes. Which one would you work in all day? 👀
13 voted
72%
5 voted
28%
18 votes
Closed
Keeping the research plan and cited results side by side is a useful way to make the agent’s work easier to inspect. I’d lean toward light mode for reading the table in daylight. How do you distinguish a verified source from one that still needs review?