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/core/tests/suite/auto_review.rs

@@ -261,6 +261,7 @@ fn remote_model_with_auto_review_override(slug: &str, review_model: &str) -> Mod
         supports_search_tool: false,
         use_responses_lite: false,
         auto_review_model_override: Some(review_model.to_string()),
+        model_specialty: None,
         tool_mode: None,
         multi_agent_version: None,
         priority: 1,

codex-rs/core/tests/suite/model_switching.rs

@@ -90,6 +90,7 @@ fn test_model_info(
         supports_search_tool: false,
         use_responses_lite: false,
         auto_review_model_override: None,
+        model_specialty: None,
         tool_mode: None,
         multi_agent_version: None,
         priority: 1,
@@ -954,6 +955,7 @@ async fn model_switch_to_smaller_model_updates_token_context_window() -> Result<
         supports_search_tool: false,
         use_responses_lite: false,
         auto_review_model_override: None,
+        model_specialty: None,
         tool_mode: None,
         multi_agent_version: None,
         priority: 1,

codex-rs/core/tests/suite/models_cache_ttl.rs

@@ -534,6 +534,7 @@ fn test_remote_model(slug: &str, priority: i32) -> ModelInfo {
         supports_search_tool: false,
         use_responses_lite: false,
         auto_review_model_override: None,
+        model_specialty: None,
         tool_mode: None,
         multi_agent_version: None,
     }

codex-rs/core/tests/suite/personality.rs

@@ -669,6 +669,7 @@ async fn remote_model_friendly_personality_instructions_with_feature() -> anyhow
         supports_search_tool: false,
         use_responses_lite: false,
         auto_review_model_override: None,
+        model_specialty: None,
         tool_mode: None,
         multi_agent_version: None,
     };
@@ -790,6 +791,7 @@ async fn user_turn_personality_remote_model_template_includes_update_message() -
         supports_search_tool: false,
         use_responses_lite: false,
         auto_review_model_override: None,
+        model_specialty: None,
         tool_mode: None,
         multi_agent_version: None,
     };

codex-rs/core/tests/suite/remote_models.rs

@@ -528,6 +528,7 @@ async fn remote_models_remote_model_uses_unified_exec() -> Result<()> {
         supports_search_tool: false,
         use_responses_lite: false,
         auto_review_model_override: None,
+        model_specialty: None,
         tool_mode: None,
         multi_agent_version: None,
         priority: 1,
@@ -782,6 +783,7 @@ async fn remote_models_apply_legacy_instructions() -> Result<()> {
         supports_search_tool: false,
         use_responses_lite: false,
         auto_review_model_override: None,
+        model_specialty: None,
         tool_mode: None,
         multi_agent_version: None,
         priority: 1,
@@ -1365,6 +1367,7 @@ fn test_remote_model_with_policy(
         supports_search_tool: false,
         use_responses_lite: false,
         auto_review_model_override: None,
+        model_specialty: None,
         tool_mode: None,
         multi_agent_version: None,
         priority,

codex-rs/core/tests/suite/rmcp_client.rs

@@ -2231,6 +2231,7 @@ async fn stdio_image_responses_are_sanitized_for_text_only_model() -> anyhow::Re
                 supports_search_tool: false,
                 use_responses_lite: false,
                 auto_review_model_override: None,
+                model_specialty: None,
                 tool_mode: None,
                 multi_agent_version: None,
             }],

codex-rs/core/tests/suite/spawn_agent_description.rs

@@ -71,6 +71,7 @@ fn test_model_info(
         supports_search_tool: false,
         use_responses_lite: false,
         auto_review_model_override: None,
+        model_specialty: None,
         tool_mode: None,
         multi_agent_version: None,
         priority: 1,

codex-rs/core/tests/suite/view_image.rs

@@ -1363,6 +1363,7 @@ async fn view_image_tool_returns_unsupported_message_for_text_only_model() -> an
         supports_search_tool: false,
         use_responses_lite: false,
         auto_review_model_override: None,
+        model_specialty: None,
         tool_mode: None,
         multi_agent_version: None,
         priority: 1,