Would you be open to a PR for Codex multi-account support?

[monterrr]

Hi! I’ve been using a private variant of CodexBar and ended up implementing Codex multi-account support for my own use.

Before I spend time polishing it for upstream submission, I wanted to ask whether this is something you’d even be interested in accepting as a PR.

Problem / motivation

Right now, if someone uses more than one Codex/OpenAI account or workspace, the app experience is pretty limited because it effectively assumes a single account/session path.

That makes it hard to:

• switch between multiple Codex accounts cleanly
• keep account/workspace identity visible
• compare or inspect usage across accounts
• avoid confusion when the active session/account changes

What my implementation currently enables

In my private repo, I implemented a working version of:

• multiple Codex accounts
• account/workspace labeling in the UI
• account-aware refresh behavior
• menu UI improvements for multi-account handling
• optional menu sorting for Codex accounts

I also did some follow-up internal cleanup to reduce maintenance pain in my private fork, but I understand that part is separate and may not be relevant upstream.

My question

Would you be open, in principle, to a PR for the feature itself if I prepared it as a clean upstream-oriented contribution?

If yes, I would plan to:

• rebase it onto a clean upstream branch
• keep the PR feature-focused
• leave out private-fork-specific cleanup unless it’s truly necessary
• split things if you’d prefer smaller pieces

If this is not a direction you want for CodexBar, no worries — I just wanted to ask before doing the cleanup work needed for an upstreamable PR.

Thanks!

[Iamadig]

oh man - would love this!

@steipete - please support this

[steipete]

I’d love to see what route you chose to tackle this, as I couldn’t come up with a great way!

[monterrr]

I opened a draft PR here so you can see the route I took: #588

[rafachavantes]

Would love this!