OpenAI 是否已通过 Codex 将临时聊天改进推送给所有用户？目前证据不足

Protecting those chats would make Codex feel safer and more reliable as a professional workspace. This improvement would help users avoid accidental disruption, especially when they are managing multiple projects, pinned chats, and long-running workflows. Thank you for considering this request. I would be happy to provide more concrete examples or screenshots from my workflow if helpful. Hi OpenAI team, I want to suggest a Codex UX improvement for users who manage complex, long-running projects with many related chats. Right now, project chats appear to be organized mostly by recent activity or internal sorting logic. That is useful in some situations, but it does not always match the real structure of a project. [...] Why this matters: Codex is becoming more than a single-task coding assistant. For many users, it is becoming a workspace where project context, execution, planning, and decisions live together. Manual project chat organization would make Codex stronger for: This

Temporary chats randomly ‘break’ (“Waiting for full response…” on all subsequent requests, along with a 404 for the /conversation completion request in the network tab.) Chats that once used a model like GPT-5.x-mini stay stuck on that model even if GPT5.5 is available; unless you force GPT5.5 by using a request editor, which isn’t quite a thing that everyone would always be willing to do. Cancelled codex OAuth2 requests sign you out of ChatGPT as well. Powered by Discourse, best viewed with JavaScript enabled [...] # Suggestions and long-pending improvements for chatgpt.com Not much to say, but a few UX tweaks are pending since long in the ChatGPT web. The search-chats would be nice if it had a hover-down preview. The timeline scrollbar on the right side of a chat conversation in ChatGPT is extremely jank, compared to the one that has been for long in the desktop app / codex. Temporary chats keep accumulating conv\_key cookies that pile up fast and clean up slower (depends on h

Comparable products already expose this pattern: Codex desktop is different from Claude Desktop: Claude has separate Chat, Cowork, and Code surfaces, while Codex desktop is primarily the coding/session interface. That means the equivalent feature should apply directly when starting a new Codex session, not only to a separate general-chat tab. Expected behavior: Why this matters: Codex sessions often have access to local files, repos, tool output, and connected app context. Users need a lightweight way to ask sensitive one-off questions or do narrow coding work without creating a durable session artifact or pulling in broader workspace context. [...] Related prior issue: #6527 requested CLI incognito mode and was closed as not planned. This request is specifically for Codex desktop UX and Temporary Chat parity with ChatGPT and Claude Desktop. ## Metadata ## Metadata ### Assignees ### Labels ### Type ### Projects ### Milestone ### Relationships ### Development ## Issue acti

either. It's pretty crazy. This is very cool. So, you're hosting it right out of the box. Exactly. And so it means that when you're doing that, you don't need to think about going and deploying uh your own custom kind of uh you know push the code to git and then you have some sort of like custom deployment and then you're going to make sure that permissions are correct on all of that. You can literally just say and again I typed it there so you can see it but I would normally just say build a site that does this and then I can go and show all of my work to my teammates and that to me is transformative again instead of sitting in silos. This information this hosted for how long? Uh I think it actually gets hosted for as long as you want. Um uh over there and uh you you don't get you don't [...] analysis. It ran some Python on it, and then I want it to create a um site for us to see this. Okay. So what I'm going to say to codeex is can you please make a site for us to visualize all of th

review. Um I want you to do a thorough security review of every single PR and give me a report. And if you find there's something necessary to fix on the spot, spin up a thread and fix it. So just something like this right and mind you this with computer use you know people talk about software factories ladies and gentlemen your software factory is codeex the only thing that was missing is a super powerful model like GPT 5.6 6 soul. So you can write an automation like this works every single morning specific time. It will review all your open PRs. It will review all security issues that you have and you just have a report. So you know we talk about being managers and managing software factories. That really wasn't possible until models like 5.6 and Fable came about. What really is the [...] I think it's really important. When you use a computer use when you use computer use on codecs, they can do it in the background. When you use it on clawed code, you'll notice there's this like kind

possible. And then for everything else, use computer use. Um how many people have used computer use by the way? Very. Oh, it's like 10% of you. What's a crazy thing that you've done? What's like the craziest thing anyone's done? Any any volunteers? Managing my home. Home lab. Yes. What's a home lab? Nice. Wow. Any other like crazy computer use stories? Got to get AGI pill and try out computer use. Um yeah, one thing I'll say is like because computers are so powerful like you get like you mentioned before there is some like safety component. I am now remembering this example where like again because the slack connector was not able to upload files. If this model is really determined, right? Uh it could be like the one wish willow. it just says, "Okay, great. Well, if I can't add a Slack uh [...] oh yeah pull request has been made it will get merged by next Thursday and like this is a crazy experience to witness right this is this actually allows us to do way more support without making