Issue Station
Sign In
Tools hereToolbox

codex - 💡(How to fix) Fix Codex CLI 0.123/0.124 regression: WebSocket fallback then stream disconnected before completion on Windows [1 comments, 2 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
openai/codex#19330•Fetched 2026-04-25 06:11:33
View on GitHub
Comments
1
Participants
2
Timeline
6
Reactions
0
Author
brolabitaliabrolabitalia
Participants
brolabitaliabrolabitaliagithub-actions[bot]github-actions[bot]
Timeline (top)
labeled ×4commented ×1unlabeled ×1

Error Message

Observed error: stream disconnected before completion: failed to send websocket request: IO error: Connessione interrotta dal software del computer host. (os error 10053) The error is Windows os error 10053, but the timing strongly correlates with the Codex CLI upgrade from 0.122.x to 0.123.x / 0.124.x.

Root Cause

The issue appears to be a regression introduced in 0.123.x, because 0.122.x was stable on the same Windows machine, terminal, project, and network.

RAW_BUFFERClick to expand / collapse

What version of Codex CLI is running?

codex-cli 0.124.0

What subscription do you have?

Using Codex CLI with ChatGPT Plus / Codex credits.

Which model were you using?

gpt-5.4 with reasoning effort xhigh The same environment was previously stable with Codex CLI 0.122.x. The transport issue started from CLI 0.123.x and persists on 0.124.x.

What platform is your computer?

Microsoft Windows NT 10.0.26200.0

What terminal emulator and version are you using (if applicable)?

Windows Terminal with PowerShell

What issue are you seeing?

Since upgrading Codex CLI from 0.122.x to 0.123.x, and still on 0.124.x, Codex sessions intermittently fail during long-running work.

The CLI first reports that it is falling back from WebSockets to HTTPS transport, then the stream disconnects before completion.

Observed error:

Falling back from WebSockets to HTTPS transport. stream disconnected before completion: failed to send websocket request: IO error: Connessione interrotta dal software del computer host. (os error 10053)

This happens during active Codex work, often after the session has been running for a while. The same machine/network was stable on CLI 0.122.x. The issue started with 0.123.x and persists on 0.124.x.

The failure interrupts the active task and can leave the session unreliable or unable to complete the response. Opening a new session sometimes helps temporarily, but the issue comes back.

What steps can reproduce the bug?

Uploaded thread: 019dbea5-9921-7112-9c3f-ae0b16f4deea

What is the expected behavior?

Codex CLI should keep the streaming connection stable and complete the task response.

If WebSocket transport fails and the CLI falls back to HTTPS, the fallback should complete reliably without disconnecting before completion. Long-running sessions should not become unreliable after the transport fallback.

Additional information

The issue appears to be a regression introduced in 0.123.x, because 0.122.x was stable on the same Windows machine, terminal, project, and network.

I also submitted feedback from inside the CLI using /feedback before opening this issue.

The error is Windows os error 10053, but the timing strongly correlates with the Codex CLI upgrade from 0.122.x to 0.123.x / 0.124.x.

extent analysis

TL;DR

Downgrade to Codex CLI version 0.122.x to potentially resolve the intermittent session failures during long-running work.

Guidance

  • Verify that the issue is indeed a regression introduced in version 0.123.x by testing with the previous stable version 0.122.x.
  • Check the system and network configuration to ensure no changes were made that could be contributing to the issue, as the same machine and network were previously stable.
  • If downgrading is not feasible, try opening a new session when the issue occurs to see if it provides a temporary workaround.
  • Monitor the issue to see if it persists across different projects and networks to rule out any environmental factors.

Example

No code snippet is provided as the issue seems to be related to the Codex CLI version and its interaction with the system and network.

Notes

The solution is based on the information provided and the fact that the issue started after upgrading from version 0.122.x. However, without further debugging or logs, it's difficult to pinpoint the exact cause.

Recommendation

Apply workaround: Downgrade to version 0.122.x, as it was the last known stable version for the user's environment. This is recommended because the issue appears to be a regression introduced in version 0.123.x, and downgrading may provide a reliable workaround until the issue is resolved in a future version.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…
#tokenizer error#prompt formatting#chain error#conversation history#tool integration

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING