Enhance model picker

[aibrahim-oai]

External (non-OpenAI) Pull Request Requirements

Before opening this Pull Request, please read the dedicated "Contributing" markdown file or your PR may be closed:
https://github.com/openai/codex/blob/main/docs/contributing.md

If your PR conforms to our contribution guidelines, replace this text with a detailed and high quality description of your changes.

Include a link to a bug report or enhancement request.

[chatgpt-codex-connector]

💡 Codex Review

codex/codex-rs/protocol/src/openai_models.rs

Lines 161 to 165 in 2a9b1e5

	upgrade: info.upgrade.as_ref().map(|upgrade_slug| ModelUpgrade {
	id: upgrade_slug.clone(),
	reasoning_effort_mapping: reasoning_effort_mapping_from_presets(
	&info.supported_reasoning_levels,
	),

Migration keeps unsupported effort for upgraded models

The ModelUpgrade built from API responses derives reasoning_effort_mapping from the source model’s supported_reasoning_levels instead of the upgrade target. If the backend now returns an upgrade to a model with fewer options (e.g., the new codex-auto-* presets that only expose a single effort), the migration flow in app.rs will carry over the user’s old effort unchanged, persisting a config.model_reasoning_effort the upgrade cannot serve and leading to failed requests after auto-migration. The mapping should be computed from the upgrade model’s supported efforts so the stored effort is clamped to what the target supports.

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".