Use the originating model when recording conversation history (#44243)

## Why

History recording used the model captured at turn start, so switching models within a turn could apply stale image preparation and tool-output truncation rules.

## What changed

Pass explicit model metadata through conversation recording. Record model responses and tool results using their originating step's settings, and capture the current model for pending input recorded during startup interruption or task completion.

## Testing

Add regression coverage for tool results across a model switch, explicit model selection for plain and annotated history, and image input retained during interruption and completion. Extend model-switching tests to verify that prepared media and item identity survive switching back to a multimodal model in live, resumed, and forked threads.

GitOrigin-RevId: 5bc750c30c334acb9b802cb80fdea01daffa84a1

codex-rs/core/src/agent/control_tests.rs

@@ -1173,6 +1173,7 @@ async fn spawn_agent_fork_from_paginated_parent_uses_model_context_prefix() {
         .session
         .record_conversation_items(
             turn_context.as_ref(),
+            turn_context.model_info(),
             &[spawn_agent_call(&parent_spawn_call_id)],
         )
         .await;
@@ -1668,7 +1669,7 @@ async fn spawn_agent_can_fork_parent_thread_history_with_sanitized_items() {
     parent_thread
         .session
         .record_conversation_items(
-            turn_context.as_ref(),
+            turn_context.as_ref(), turn_context.model_info(),
             &[
                 ResponseItem::Message {
                     id: None,
@@ -2561,6 +2562,7 @@ async fn spawn_agent_fork_flushes_parent_rollout_before_loading_history() {
         .session
         .record_conversation_items(
             turn_context.as_ref(),
+            turn_context.model_info(),
             &[
                 assistant_message("unflushed final answer", Some(MessagePhase::FinalAnswer)),
                 spawn_agent_call(&parent_spawn_call_id),
@@ -2633,6 +2635,7 @@ async fn spawn_agent_fork_last_n_turns_keeps_only_recent_turns() {
         .session
         .record_conversation_items(
             queued_turn_context.as_ref(),
+            queued_turn_context.model_info(),
             &[queued_communication.to_response_input_item().into()],
         )
         .await;
@@ -2649,6 +2652,7 @@ async fn spawn_agent_fork_last_n_turns_keeps_only_recent_turns() {
         .session
         .record_conversation_items(
             triggered_turn_context.as_ref(),
+            triggered_turn_context.model_info(),
             &[triggered_communication.to_response_input_item().into()],
         )
         .await;
@@ -2662,6 +2666,7 @@ async fn spawn_agent_fork_last_n_turns_keeps_only_recent_turns() {
         .session
         .record_conversation_items(
             spawn_turn_context.as_ref(),
+            spawn_turn_context.model_info(),
             &[spawn_agent_call(&parent_spawn_call_id)],
         )
         .await;
@@ -2774,6 +2779,7 @@ async fn spawn_agent_fork_last_n_turns_drops_parent_startup_prefix_when_under_li
         .session
         .record_conversation_items(
             startup_turn_context.as_ref(),
+            startup_turn_context.model_info(),
             &[ResponseItem::Message {
                 id: None,
                 role: "developer".to_string(),
@@ -2795,6 +2801,7 @@ async fn spawn_agent_fork_last_n_turns_drops_parent_startup_prefix_when_under_li
         .session
         .record_conversation_items(
             spawn_turn_context.as_ref(),
+            spawn_turn_context.model_info(),
             &[spawn_agent_call(&parent_spawn_call_id)],
         )
         .await;
@@ -2902,6 +2909,7 @@ async fn spawn_agent_fork_last_n_turns_strips_parent_usage_hints() {
         .session
         .record_conversation_items(
             turn_context.as_ref(),
+            turn_context.model_info(),
             &[
                 ResponseItem::Message {
                     id: None,