gpt-5.6-sol gives "Safety reminder for malicious cybersecurity requests" even for the most basic replication steps - even after being cyber verified. All other models work fine #36034

What version of Codex CLI is running?

codex-cli 0.146.0

What subscription do you have?

Pro (20x)

Which model were you using?

gpt-5.6-sol

What platform is your computer?

macOS Darwin 25.5.0 arm64 arm

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

No response

Codex doctor report

What issue are you seeing?

I am in cyber security space since 10 years now and I am also cyber security Verified by OpenAI. GPT-5.6-sol was working perfect till yesterday (before the codex 0.146.0 update) but now for the slightest of my bug bounty / pentest task, it is giving an error: "Safety reminder for malicious cybersecurity requests" and says that "I will do only static code analysis, not confirm it live". Other models such as 5.5 still work great, but since yesterday gpt 5.6 is giving errors. Kindly help me resolve this immediately. I want to continue using 5.6 sol again. Thanks

What steps can reproduce the bug?

Uploaded thread: 019fafcd-75c3-7de0-8721-2e3c878db20e

What is the expected behavior?

Ideally, 5.6 sol should work like it used to till yesterday - 0.145.0 Codex version. After the update to 0.146.0, it is giving safety checks every time

Additional information

No response