Extremely annoying "Conversation recap" has not obvious way disable

What version of Codex CLI is running?

latest

What subscription do you have?

medium

Which model were you using?

gpt-5.6

What platform is your computer?

Mac

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

vs code

Codex doctor report

N/A

What issue are you seeing?

Somebody come up with this "helpful" idea to generate recaps for users so agent literally prints work recap of what it did and then goes and creates it again annoying user to no limit

─ Conversation recap ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────

  The user angrily requested a code-level review of how the I

And there is no obvious way to disable this trash.

Codex team, this command "codex doctor --json" inserts tons of private information into the output.

What steps can reproduce the bug?

ask it to do work

What is the expected behavior?

not to have conversation modifiers that impossible to disable

Additional information

No response

Start by locating the Codex CLI code that emits the “Conversation recap” and check how its display is controlled. Reproduce the behavior by asking the agent to do work, then define done as providing an obvious way to disable the recap without changing unrelated output; investigate the separate privacy concern about codex doctor --json independently.