So, I have an app in the review queue like many others. And it will probably be at least a week and maybe more before they even start looking at them. And presumably there is a backlog. So, waiting…
During that period, are there rules or guidance on if we are allowed to tweak the app, fix bugs we find, etc? Or is that forbidden / should be relegated only to a staging server and then submitted as an update later? Or some things are fine but others (tool definitions, CSP) are not fine?
I can’t find anything on this topic in the docs.
2 Likes
Gonna try re-upping this since people are back now.
@OpenAI_Support Do you know?
1 Like
Also interested in this - many people also seemed to be confused about the country in/exclusion in the UI.
1 Like
From the OpenAI Help Center article “Submitting apps to the ChatGPT app directory”:
- You don’t need to take extra steps while review is in progress unless OpenAI contacts you.
- If you want the version in-review to change, the supported path is: Cancel review → resubmit (and describe the changes).
- Updates (after approval) also go through resubmission + review, and OpenAI recommends only one version in review at a time.
So: I’d keep the submitted version stable (esp. tool defs/auth/CSP/privacy), keep iterating on staging, and only “swap” what reviewers see by cancel+resubmit when you’re ready.
This matches up with my vague head-canon on what we’re supposed to do here, absent anything more explicit.