Record sandbox backends in command execution analytics (#47568)

## Why

Command execution analytics lack the observed sandbox backend. Older execution servers may omit this information, so an unknown backend must remain distinguishable from an explicitly unsandboxed process.

## What changed

- Carry optional `sandbox_type` metadata through command execution items and emit its metric tag as `sandbox_backend` in analytics.
- Propagate the observed backend through unified execution completion and failure events, including asynchronous exits.
- Preserve missing backend information as `None` and mark executed user shell commands as `Some(SandboxType::None)`.
- Exclude the metadata from serialized command items and generated schemas and TypeScript types.

## Testing

Extend tests for missing and explicit executor backends, backend propagation on failed execution, and analytics payloads, including platform-specific backend tags in app-server integration coverage.

GitOrigin-RevId: 178fe46824a6194910802ec02168ed40ee0b2019
This commit is contained in:
iceweasel-oai
2026-09-23 15:05:59 +00:00
committed by copyberry
parent d6c4c6aea4
commit 6f6af0fce4
29 changed files with 113 additions and 21 deletions
@@ -49,6 +49,7 @@ use codex_app_server_protocol::ServerNotification;
use codex_app_server_protocol::ThreadItem;
use codex_app_server_protocol::TurnStatus as AppServerTurnStatus;
use codex_protocol::protocol::ThreadSource;
use codex_protocol::sandbox::SandboxType;
use codex_utils_absolute_path::test_support::PathBufExt;
use codex_utils_absolute_path::test_support::test_path_buf;
use pretty_assertions::assert_eq;
@@ -61,12 +62,14 @@ fn sample_command_execution_item_with_actions(
command_actions: Vec<CommandAction>,
plugin_id: Option<&str>,
script_path: Option<&str>,
sandbox_type: Option<SandboxType>,
) -> ThreadItem {
let mut item = sample_command_execution_item(status, exit_code, duration_ms);
let ThreadItem::CommandExecution {
command_actions: item_command_actions,
plugin_id: item_plugin_id,
script_path: item_script_path,
sandbox_type: item_sandbox_type,
..
} = &mut item
else {
@@ -75,6 +78,7 @@ fn sample_command_execution_item_with_actions(
*item_command_actions = command_actions;
*item_plugin_id = plugin_id.map(str::to_string);
*item_script_path = script_path.map(str::to_string);
*item_sandbox_type = sandbox_type;
item
}
@@ -96,6 +100,7 @@ fn command_execution_event_serializes_expected_shape() {
let event = TrackEventRequest::CommandExecution(CodexCommandExecutionEventRequest {
event_type: "codex_command_execution_event",
event_params: CodexCommandExecutionEventParams {
sandbox_backend: None,
model_slug: None,
reasoning_effort: None,
base: CodexToolItemEventBase {
@@ -155,6 +160,7 @@ fn command_execution_event_serializes_expected_shape() {
let mut expected = json!({
"event_type": "codex_command_execution_event",
"event_params": {
"sandbox_backend": null,
"model_slug": null,
"reasoning_effort": null,
"thread_id": "thread-1",
@@ -291,6 +297,7 @@ async fn item_lifecycle_notifications_publish_command_execution_event() {
],
Some("sample@openai-curated"),
Some("scripts/run.py"),
Some(SandboxType::WindowsMxc),
),
},
))),
@@ -301,6 +308,7 @@ async fn item_lifecycle_notifications_publish_command_execution_event() {
let payload = serde_json::to_value(&events).expect("serialize events");
assert_eq!(payload.as_array().expect("events array").len(), 1);
assert_eq!(payload[0]["event_params"]["model_slug"], "invoking-model");
assert_eq!(payload[0]["event_params"]["sandbox_backend"], "windows_mxc");
assert_eq!(payload[0]["event_params"]["reasoning_effort"], "max");
assert_eq!(payload[0]["event_type"], "codex_command_execution_event");
assert_eq!(payload[0]["event_params"]["thread_id"], "thread-1");
+1
View File
@@ -647,6 +647,7 @@ pub(super) fn sample_command_execution_item_with_id(
) -> ThreadItem {
ThreadItem::CommandExecution {
model_context: None,
sandbox_type: None,
id: id.to_string(),
plugin_id: None,
script_path: None,