Codex App is not working even though I am logged in with Enterprise account

Resolved 💬 2 comments Opened May 8, 2026 by kkambhat Closed May 12, 2026

What version of the Codex App are you using (From “About Codex” dialog)?

Version 26.506.31004 (2604)

What subscription do you have?

Enterprise Subscription from Cisco

What platform is your computer?

MAC OS Tahoe 26.4.1 (25E253)

What issue are you seeing?

<img width="929" height="527" alt="Image" src="https://github.com/user-attachments/assets/eb72b7d6-302e-4821-99aa-a11fc740a572" />
I am logged into the Cisco Enterprise Account through ChatGPT. After logging in when i try to change the model or try to create a project or chat, nothing Works and I get the message as shown above.

I observe the same in the CLI too

kkambhat@KKAMBHAT-M-DWY2 NetSenseAI % codex --yolo
Error loading configuration: Failed to load cloud requirements (workspace-managed policies).

The CLI version is
kkambhat@KKAMBHAT-M-DWY2 NetSenseAI % codex --version
codex-cli 0.129.

I have deleted and re-installed CLI and the Codex App and re-tried but the same error.

What steps can reproduce the bug?

Pretty straightforward. Install and try and it bails out immediately. Pretty bad issue. No problems logging into the account.

What is the expected behavior?

The expected behavior is to be able to
(1) Start the CLI after you login as well as be able to use the Desktop app.
(2) Be able to communicate with the workspace immediately

Additional information

_No response_

View original on GitHub ↗

2 Comments

kkambhat · 3 months ago

Ok an update here. I had to completely uninstall codex CLI and APP completely from my MAC . Then install an older version of the codex app 0.124.0 and then it started working again !

rm -rf ~/.codex
rm -rf ~/Library/Application\ Support/Codex
rm -rf ~/Library/Caches/Codex
rm -rf ~/Library/Caches/com.openai.codex*
rm -rf ~/Library/Preferences/com.openai.codex*
rm -rf ~/Library/Logs/Codex

ls /Applications | grep -i codex
# If found:
rm -rf  /Applications/Codex.app

Then I did an update from within the CLI and it conitnues to work fine ! So definitely codex install directly to 0.130 is not working.

kkambhat@KKAMBHAT-M-DWY2 % codex --yolo 

╭─────────────────────────────────────────────────╮
│ ✨ Update available! 0.124.0 -> 0.130.0         │
│ Run npm install -g @openai/codex to update.     │
│                                                 │
│ See full release notes:                         │
│ https://github.com/openai/codex/releases/latest │
╰─────────────────────────────────────────────────╯

Once you exit and log back in again it does not work anymore.

Even in version 0.124.0 I see this wierd error "⚠ Falling back from WebSockets to HTTPS transport. stream disconnected before completion: invalid peer certificate: UnknownIssuer"

kkambhat · 3 months ago

This is how I resolved the issue for systems connected to Corporate VPNs which is my case and trying to OAuth to Codex.

Export the corporate CA bundle and tell Codex to use it. For MAC:

Step 1:
mkdir -p ~/.ssl
Step 2:
security find-certificate -a -p /Library/Keychains/System.keychain > ~/.ssl/corp-ca-bundle.pem
Step 3:
security find-certificate -a -p /System/Library/Keychains/SystemRootCertificates.keychain >> ~/.ssl/corp-ca-bundle.pem

Add to your shell profile (~/.zshrc or ~/.bashrc)

export SSL_CERT_FILE=~/.ssl/corp-ca-bundle.pem

Restart codex CLI and codex APP

Step 1: Exit the codex CLI and Codex App

Step 2: In the terminal do a exec ${SHELL}

Step 3: In the terminal, do
a. codex login
b. codex

Step 4: Start the Codex Desktop app.