Add safer TUI defaults for cyber models (#37055)

## What changed

- Propagate the optional `modelSpecialty` field through the model catalog and
  app-server `model/list` response.
- When a newly selected model has the `cyber` specialty, default an active TUI
  thread to workspace-write permissions with on-request approval. Use auto
  review when available, otherwise keep the user as reviewer, while respecting
  configured permission requirements.
- Show a notice when auto review is applied and strengthen the full-access
  warning for cyber models.
- Preserve explicitly selected permissions when only reasoning settings change.

## Testing

- Cover auto-review defaults, requirement-aware fallback behavior, permission
  preservation across reasoning changes, and the new warning notice.

GitOrigin-RevId: 8f25b0082d3051d9ff91f7355921236d0f82f758

codex-rs/app-server/tests/suite/v2/model_list.rs

@@ -43,6 +43,7 @@ fn model_from_preset(preset: &ModelPreset) -> Model {
         availability_nux: preset.availability_nux.clone().map(Into::into),
         display_name: preset.display_name.clone(),
         description: preset.description.clone(),
+        model_specialty: preset.model_specialty.clone(),
         hidden: !preset.show_in_picker,
         supported_reasoning_efforts: preset
             .supported_reasoning_efforts
@@ -156,6 +157,7 @@ async fn list_models_uses_chatgpt_remote_catalog_as_source_of_truth() -> Result<
         "slug": "chatgpt-remote-only",
         "display_name": "ChatGPT Remote Only",
         "description": "Remote-only model for app-server model/list coverage",
+        "model_specialty": "cyber",
         "default_reasoning_level": "max",
         "supported_reasoning_levels": [
             {"effort": "max", "description": "Maximum"},