(pty “Device not configured”)
Resolved 💬 6 comments Opened Nov 24, 2025 by saulrichardson Closed Nov 26, 2025
💡 Likely answer: A maintainer (github-actions[bot], contributor)
responded on this thread — see the highlighted reply below.
What version of Codex is running?
0.63.0
What subscription do you have?
Pro
Which model were you using?
gpt-5.1-codex-max
What platform is your computer?
Mac
What issue are you seeing?
sometimes sessions cant execute anything and codex gives some version of the following:
command execution here is blocked
(pty “Device not configured”)
What steps can reproduce the bug?
Not sure. maybe its long winded sessions or something in my config. I do have some of the experimental features turned on.
What is the expected behavior?
_No response_
Additional information
fixed when session is stoped and resume
6 Comments
Potential duplicates detected. Please review them and close your issue if it is a duplicate.
Powered by Codex Action
Hi @saulrichardson, thanks for filling this bug. Could you tell me exactly which experimental features are turned on? Does it rather happen on rather "long" sessions where you tend to consume a lot of tokens and codex spawns a lot of commands?
I opened https://github.com/openai/codex/issues/7245 before seeing this issue, can confirm that this is
unified_execleaking tty file descriptors.I have unified exec on. It usually happens after medium ish sessions.
config.toml:
web_search_request = true
unified_exec = true
streamable_shell = true
apply_patch_freeform = true
rmcp_client = true
You can remove streamable_shell, this option doesn't exist anymore. Delete the unified_exec line and restart codex to fix it.
You can use "run with tmux" as a workaround.
Thanks @steipete, this issue should be fixed with #7239 with the next release same as for #7245 :)