I want ChatGPT and Codex to work together more directly.
ChatGPT should be able to turn a long design conversation into a Codex task, Codex should implement it, and ChatGPT should explain the result back to the user in a beginner-friendly way.
The user should not have to manually copy context between ChatGPT and Codex. The user should mainly approve or reject the risky parts.
Even a shared working memo for task rules, change history, and current implementation context would reduce a lot of manual copying between ChatGPT and Codex.
Maybe I am useing it wrong (i am aware of codex) but I have been buildign a large stock/orders/billing/job handelign suit with ChatGPT for my work, I found the canvas suitable for testing then just deploy the files i get, i never had to click into codex though. we are on patch 20 and still going strong, I have maxed out several conversations though, and i am handing the files off to the next chat, making 1 module at a time. To handover i ask for a detailed handover text, if i get it in txt file it is currently ~40Kb, i seen others on memory discussion estimating chats memory is <100Kb for any level of user, so i have asked for chat to use a project txt file it stores, as a memory file from now on. not sure if the project folder translate over to codex, but all chats in a project should be able to access the “project files” so hopefully memory issue solved, but i would be concerned what u asked it to hand over to codex may have been too large for its memory, depending on the complexity of your project, a chat can hold a novel, but it memory is only fit for a few pages as i understand
Thanks, that’s helpful. I’ll probably use actual project files such as AGENTS.md or handover.md rather than relying on chat memory.
My case is a little different because most of the work is currently local, and I’m planning to commit cleaned Markdown posts to GitHub in batches. So I’m still trying to figure out the best way to keep ChatGPT/Codex/project context aligned across local files, project files, and the repo.
Hey @HanakoAbsed, I agree that can be convenient.
At the moment the conversations between ChatGPT and Codex remain separate and this behavior is expected at this time. Your topic and feature request comes up regularly. Also refer this documentation for more info.
Props to @OctoMunky sharing the workaround.
- Avinash