Ctrl+T transcript overlay truncates very long single lines instead of wrapping to viewport width

What version of Codex is running?

codex-cli 0.63.0

What subscription do you have?

API

Which model were you using?

gpt-5.1-codex

What platform is your computer?

Linux 6.6.87.2-microsoft-standard-WSL2 x86_64 x86_64 (Windows 11 host with WSL2 Ubuntu)

What issue are you seeing?

In the Ctrl+T transcript overlay, very long single lines are rendered as a single on‑screen row and simply cut off at the right edge of the terminal, instead of being wrapped to the viewport width.

This makes it impossible to inspect the full content of very long tool outputs or command outputs in the transcript view: you only see the leftmost part of the line, and there is no horizontal scrolling or soft-wrapping applied there.

In my case, I reproduced this with a Python tool call that prints a single line of about 1,000 characters. In the normal TUI view, the output is visible, but when I open the Ctrl+T transcript overlay the same output appears as only one row and is truncated to the terminal width.

What steps can reproduce the bug?

1. Start codex in a terminal.

1. In the composer, send a message that makes Codex run a tool/exec command which prints a single extremely long line (around 1,000 characters). For example, ask it something like:

Execute a tool call, using Python to print a very long single line (1000 characters). I need to test if this project correctly displays long text.

Concretely, Codex ends up running something like:

• python -c "print('X'*1000, end='')"

1. Wait for the tool call / command to finish.

1. Optionally scroll the normal TUI view and confirm that the output is present (it may be wrapped or partially visible there, but it is at least accessible).

<img width="1729" height="920" alt="Image" src="https://github.com/user-attachments/assets/abf33607-d8d4-4512-91a2-1b3674cd0fed" />

1. Press Ctrl+T to open the transcript overlay.

1. In the transcript overlay, locate the tool/exec output from step 2.

• You will see that the 1,000-character line is rendered as a single row and is cut off at the viewport width.
• There is no soft wrapping within the transcript overlay and no horizontal scroll, so the majority of the line cannot be viewed.

<img width="1731" height="925" alt="Image" src="https://github.com/user-attachments/assets/6f5fc042-ecb8-48a6-b2c0-7d78eae9286d" />

What is the expected behavior?

I expect the Ctrl+T transcript overlay to be width-aware and to soft-wrap long logical lines to the current viewport width, just like other TUI components that render history lines.

Concretely:

• A very long single line (e.g. 1,000 characters of Python output) should wrap into multiple on‑screen rows in the transcript overlay so that the full content is readable with normal vertical scrolling.
• The effective height of the transcript for that entry should reflect the wrapped height, so that paging/scrolling behaves consistently.
• Alternatively, if hard wrapping is not desired, there should be some way to inspect the full line (e.g. horizontal scrolling or a dedicated view), but the current behavior of silently truncating at the right edge gives no way to see the full output.

Additional information

_No response_