Generate session name based on full session contents rather than first prompt
Resolved 💬 9 comments Opened Jan 26, 2026 by happyf-weallareeuropean Closed Mar 14, 2026
💡 Likely answer: A maintainer (github-actions[bot], contributor)
responded on this thread — see the highlighted reply below.
What feature would you like to see?
A.Search should index the entire session history, not only the first user prompt of each session. and without entiring the session by keydown enter key to jump to first/next match keyword in that selected session , you even can scroll it, shift enter to backward jump, cmd enter to enter the resume, jk as up and down to browse
B.default to search entire codex resume. optionally being limited in the project root.
Additional information
_No response_
9 Comments
Potential duplicates detected. Please review them and close your issue if it is a duplicate.
Powered by Codex Action
Which search feature are you referring to here? I'm not aware of any
resume search. Or is this a new feature that you're proposing that we add?if you know there's a things call /resume
currently when you enter the /resume you should see it say 'type to search'. which this part i descript it as resume search
Ah, I understand what you mean now. I was confused by your use of the term "search" since there's no search operation here.
I think what you're saying is that you would like to see better session names. I've updated the title of the feature request to make this clearer.
@etraut-openai much of the pain of resuming conversations can be ameliorated by middle-truncating the Conversation field in the /resume screen, similar to how the Mac Finder middle-truncates filenames.
You can see how this solution would improve from the current /resume screen implementation in the issue I filed: https://github.com/openai/codex/issues/8239
An additional benefit to this solution is that the enhancement is much less involved than working on a new session naming feature (https://github.com/openai/codex/issues/4163).
@miraclebakelaser, yes I agree that would be easier to implement. I'll note, however, that #8239 hasn't received any upvotes, so it's not resonating with other Codex users.
@etraut-openai fair enough
This feature request hasn't received any upvotes, so I'm closing it.