[codex] Preserve raw code-mode exec output by default (#23564)

## Why
Code mode can use nested unified exec calls as data sources. When those
calls omit `max_output_tokens`, code mode should receive raw command
output so the script can parse or summarize it itself. When code mode
does provide `max_output_tokens`, that explicit nested budget should be
respected, including values above the default unified exec limit, rather
than being capped before code mode sees the result.

## What
- Preserve direct unified exec truncation behavior, while letting
code-mode exec/write_stdin keep `max_output_tokens` as `None` unless
explicitly supplied.
- Make code-mode tool results use raw output when no explicit limit is
present, and use the explicit nested limit directly when one is
specified.
- Refactor unified exec output formatting so `truncated_output` takes
the caller-selected token budget.
- Add e2e integration coverage for explicit nested exec limits, omitted
nested exec limits, outer exec limit propagation, omitted-limit outputs
that exceed both the default and a small truncation policy, explicit
nested limits above those caps, and high explicit limits that still
compact larger command output.
- Reuse the code-mode turn setup helper while directly asserting the
exact exec output item in each test.

## Testing
- `just fmt`
- `git diff --check`
- Not run locally per repo guidance; CI should validate the e2e
integration tests.

codex-rs/core/src/unified_exec/mod.rs

@@ -31,6 +31,7 @@ use codex_exec_server::Environment;
 use codex_network_proxy::NetworkProxy;
 use codex_protocol::models::AdditionalPermissionProfile;
 use codex_utils_absolute_path::AbsolutePathBuf;
+use codex_utils_output_truncation::TruncationPolicy;
 use rand::Rng;
 use rand::rng;
 use tokio::sync::Mutex;
@@ -111,6 +112,7 @@ pub(crate) struct WriteStdinRequest<'a> {
     pub input: &'a str,
     pub yield_time_ms: u64,
     pub max_output_tokens: Option<usize>,
+    pub truncation_policy: TruncationPolicy,
 }
 
 #[derive(Default)]

codex-rs/core/src/unified_exec/mod_tests.rs

@@ -10,6 +10,7 @@ use crate::tools::context::ExecCommandToolOutput;
 use crate::unified_exec::WriteStdinRequest;
 use crate::unified_exec::process::OutputHandles;
 use codex_sandboxing::SandboxType;
+use codex_utils_output_truncation::TruncationPolicy;
 use codex_utils_output_truncation::approx_token_count;
 use core_test_support::get_remote_test_env;
 use core_test_support::skip_if_sandbox;
@@ -162,6 +163,7 @@ async fn exec_command_with_tty(
         chunk_id: generate_chunk_id(),
         wall_time,
         raw_output: collected,
+        truncation_policy: turn.truncation_policy,
         max_output_tokens: None,
         process_id: response_process_id,
         exit_code,
@@ -195,6 +197,7 @@ async fn write_stdin(
             input,
             yield_time_ms,
             max_output_tokens: None,
+            truncation_policy: TruncationPolicy::Tokens(10_000),
         })
         .await
 }
@@ -260,7 +263,9 @@ async fn unified_exec_persists_across_requests() -> anyhow::Result<()> {
     )
     .await?;
     assert!(
-        out_2.truncated_output().contains("codex"),
+        out_2
+            .truncated_output(DEFAULT_MAX_OUTPUT_TOKENS)
+            .contains("codex"),
         "expected environment variable output"
     );
 
@@ -301,7 +306,9 @@ async fn multi_unified_exec_sessions() -> anyhow::Result<()> {
         "short command should not report a process id if it exits quickly"
     );
     assert!(
-        !out_2.truncated_output().contains("codex"),
+        !out_2
+            .truncated_output(DEFAULT_MAX_OUTPUT_TOKENS)
+            .contains("codex"),
         "short command should run in a fresh shell"
     );
 
@@ -313,7 +320,9 @@ async fn multi_unified_exec_sessions() -> anyhow::Result<()> {
     )
     .await?;
     assert!(
-        out_3.truncated_output().contains("codex"),
+        out_3
+            .truncated_output(DEFAULT_MAX_OUTPUT_TOKENS)
+            .contains("codex"),
         "session should preserve state"
     );
 
@@ -350,7 +359,9 @@ async fn unified_exec_timeouts() -> anyhow::Result<()> {
     )
     .await?;
     assert!(
-        !out_2.truncated_output().contains(TEST_VAR_VALUE),
+        !out_2
+            .truncated_output(DEFAULT_MAX_OUTPUT_TOKENS)
+            .contains(TEST_VAR_VALUE),
         "timeout too short should yield incomplete output"
     );
 
@@ -359,7 +370,9 @@ async fn unified_exec_timeouts() -> anyhow::Result<()> {
     let out_3 = write_stdin(&session, process_id, "", /*yield_time_ms*/ 100).await?;
 
     assert!(
-        out_3.truncated_output().contains(TEST_VAR_VALUE),
+        out_3
+            .truncated_output(DEFAULT_MAX_OUTPUT_TOKENS)
+            .contains(TEST_VAR_VALUE),
         "subsequent poll should retrieve output"
     );
 
@@ -394,7 +407,9 @@ async fn unified_exec_pause_blocks_yield_timeout() -> anyhow::Result<()> {
         "pause should block the unified exec yield timeout"
     );
     assert!(
-        response.truncated_output().contains("unified-exec-done"),
+        response
+            .truncated_output(DEFAULT_MAX_OUTPUT_TOKENS)
+            .contains("unified-exec-done"),
         "exec_command should wait for output after the pause lifts"
     );
     assert!(
@@ -420,7 +435,11 @@ async fn requests_with_large_timeout_are_capped() -> anyhow::Result<()> {
     .await?;
 
     assert!(result.process_id.is_some());
-    assert!(result.truncated_output().contains("codex"));
+    assert!(
+        result
+            .truncated_output(DEFAULT_MAX_OUTPUT_TOKENS)
+            .contains("codex")
+    );
 
     Ok(())
 }
@@ -442,7 +461,11 @@ async fn completed_commands_do_not_persist_sessions() -> anyhow::Result<()> {
         result.process_id.is_some(),
         "completed command should report a process id"
     );
-    assert!(result.truncated_output().contains("codex"));
+    assert!(
+        result
+            .truncated_output(DEFAULT_MAX_OUTPUT_TOKENS)
+            .contains("codex")
+    );
 
     assert!(
         session

codex-rs/core/src/unified_exec/process_manager.rs

@@ -581,6 +581,7 @@ impl UnifiedExecProcessManager {
             chunk_id,
             wall_time,
             raw_output: collected,
+            truncation_policy: context.turn.truncation_policy,
             max_output_tokens: request.max_output_tokens,
             process_id: response_process_id,
             exit_code,
@@ -725,6 +726,7 @@ impl UnifiedExecProcessManager {
             chunk_id,
             wall_time,
             raw_output: collected,
+            truncation_policy: request.truncation_policy,
             max_output_tokens: request.max_output_tokens,
             process_id,
             exit_code,