Hi, thanks for putting this together.
From the README, it looks like this runner auto-detects and invokes the native Windows codex.exe, which aligns with the OpenAI docs for running Codex natively (using the experimental Windows sandbox).
I wanted to ask:
Is this implementation intentionally limited to the native Windows Codex CLI?
Would it be possible (or reasonable) to use Codex via WSL instead (for example, invoking codex inside WSL) since it usually gives get better results?
I understand this may be out of scope for the project, but I wanted to check whether WSL support was considered or if there are known blockers.
Thanks again!
Hi, thanks for putting this together.
From the README, it looks like this runner auto-detects and invokes the native Windows codex.exe, which aligns with the OpenAI docs for running Codex natively (using the experimental Windows sandbox).
I wanted to ask:
Is this implementation intentionally limited to the native Windows Codex CLI?
Would it be possible (or reasonable) to use Codex via WSL instead (for example, invoking codex inside WSL) since it usually gives get better results?
I understand this may be out of scope for the project, but I wanted to check whether WSL support was considered or if there are known blockers.
Thanks again!