[core] add optional status_code to error events (#6865)

We want to better uncover error status code for clients. Add an optional
status_code to error events (thread error, error, stream error) so app
server could uncover the status code from the client side later.

in event log:
```
< {
<   "method": "codex/event/stream_error",
<   "params": {
<     "conversationId": "019a9a32-f576-7292-9711-8e57e8063536",
<     "id": "0",
<     "msg": {
<       "message": "Reconnecting... 5/5",
<       "status_code": 401,
<       "type": "stream_error"
<     }
<   }
< }
< {
<   "method": "codex/event/error",
<   "params": {
<     "conversationId": "019a9a32-f576-7292-9711-8e57e8063536",
<     "id": "0",
<     "msg": {
<       "message": "exceeded retry limit, last status: 401 Unauthorized, request id: 9a0cb03a485067f7-SJC",
<       "status_code": 401,
<       "type": "error"
<     }
<   }
< }
```
This commit is contained in:
Celia Chen
2025-11-19 11:51:21 -08:00
committed by GitHub
Unverified
parent 20982d5c6a
commit c2ec477d93
10 changed files with 101 additions and 24 deletions
+8 -5
View File
@@ -66,6 +66,7 @@ use crate::context_manager::ContextManager;
use crate::environment_context::EnvironmentContext;
use crate::error::CodexErr;
use crate::error::Result as CodexResult;
use crate::error::http_status_code_value;
#[cfg(test)]
use crate::exec::StreamOutput;
use crate::mcp::auth::compute_auth_statuses;
@@ -79,7 +80,6 @@ use crate::protocol::ApplyPatchApprovalRequestEvent;
use crate::protocol::AskForApproval;
use crate::protocol::BackgroundEventEvent;
use crate::protocol::DeprecationNoticeEvent;
use crate::protocol::ErrorEvent;
use crate::protocol::Event;
use crate::protocol::EventMsg;
use crate::protocol::ExecApprovalRequestEvent;
@@ -133,6 +133,7 @@ use codex_protocol::user_input::UserInput;
use codex_utils_readiness::Readiness;
use codex_utils_readiness::ReadinessFlag;
use codex_utils_tokenizer::warm_model_cache;
use reqwest::StatusCode;
/// The high-level interface to the Codex system.
/// It operates as a queue pair where you send submissions and receive events.
@@ -1186,9 +1187,11 @@ impl Session {
&self,
turn_context: &TurnContext,
message: impl Into<String>,
http_status_code: Option<StatusCode>,
) {
let event = EventMsg::StreamError(StreamErrorEvent {
message: message.into(),
http_status_code: http_status_code_value(http_status_code),
});
self.send_event(turn_context, event).await;
}
@@ -1680,6 +1683,7 @@ mod handlers {
id: sub_id.clone(),
msg: EventMsg::Error(ErrorEvent {
message: "Failed to shutdown rollout recorder".to_string(),
http_status_code: None,
}),
};
sess.send_event_raw(event).await;
@@ -1933,10 +1937,8 @@ pub(crate) async fn run_task(
}
Err(e) => {
info!("Turn error: {e:#}");
let event = EventMsg::Error(ErrorEvent {
message: e.to_string(),
});
sess.send_event(&turn_context, event).await;
sess.send_event(&turn_context, EventMsg::Error(e.to_error_event(None)))
.await;
// let the user continue the conversation
break;
}
@@ -2060,6 +2062,7 @@ async fn run_turn(
sess.notify_stream_error(
&turn_context,
format!("Reconnecting... {retries}/{max_retries}"),
e.http_status_code(),
)
.await;
+5 -9
View File
@@ -10,7 +10,6 @@ use crate::error::Result as CodexResult;
use crate::features::Feature;
use crate::protocol::AgentMessageEvent;
use crate::protocol::CompactedItem;
use crate::protocol::ErrorEvent;
use crate::protocol::EventMsg;
use crate::protocol::TaskStartedEvent;
use crate::protocol::TurnContextItem;
@@ -128,10 +127,8 @@ async fn run_compact_task_inner(
continue;
}
sess.set_total_tokens_full(turn_context.as_ref()).await;
let event = EventMsg::Error(ErrorEvent {
message: e.to_string(),
});
sess.send_event(&turn_context, event).await;
sess.send_event(&turn_context, EventMsg::Error(e.to_error_event(None)))
.await;
return;
}
Err(e) => {
@@ -141,15 +138,14 @@ async fn run_compact_task_inner(
sess.notify_stream_error(
turn_context.as_ref(),
format!("Reconnecting... {retries}/{max_retries}"),
e.http_status_code(),
)
.await;
tokio::time::sleep(delay).await;
continue;
} else {
let event = EventMsg::Error(ErrorEvent {
message: e.to_string(),
});
sess.send_event(&turn_context, event).await;
sess.send_event(&turn_context, EventMsg::Error(e.to_error_event(None)))
.await;
return;
}
}
+2 -5
View File
@@ -6,7 +6,6 @@ use crate::codex::TurnContext;
use crate::error::Result as CodexResult;
use crate::protocol::AgentMessageEvent;
use crate::protocol::CompactedItem;
use crate::protocol::ErrorEvent;
use crate::protocol::EventMsg;
use crate::protocol::RolloutItem;
use crate::protocol::TaskStartedEvent;
@@ -30,10 +29,8 @@ pub(crate) async fn run_remote_compact_task(sess: Arc<Session>, turn_context: Ar
async fn run_remote_compact_task_inner(sess: &Arc<Session>, turn_context: &Arc<TurnContext>) {
if let Err(err) = run_remote_compact_task_inner_impl(sess, turn_context).await {
let event = EventMsg::Error(ErrorEvent {
message: format!("Error running remote compact task: {err}"),
});
sess.send_event(turn_context, event).await;
let event = err.to_error_event(Some("Error running remote compact task".to_string()));
sess.send_event(turn_context, EventMsg::Error(event)).await;
}
}
+71
View File
@@ -10,6 +10,7 @@ use chrono::Local;
use chrono::Utc;
use codex_async_utils::CancelErr;
use codex_protocol::ConversationId;
use codex_protocol::protocol::ErrorEvent;
use codex_protocol::protocol::RateLimitSnapshot;
use reqwest::StatusCode;
use serde_json;
@@ -430,6 +431,37 @@ impl CodexErr {
pub fn downcast_ref<T: std::any::Any>(&self) -> Option<&T> {
(self as &dyn std::any::Any).downcast_ref::<T>()
}
pub fn http_status_code(&self) -> Option<StatusCode> {
match self {
CodexErr::UnexpectedStatus(err) => Some(err.status),
CodexErr::RetryLimit(err) => Some(err.status),
CodexErr::UsageLimitReached(_) | CodexErr::UsageNotIncluded => {
Some(StatusCode::TOO_MANY_REQUESTS)
}
CodexErr::InternalServerError => Some(StatusCode::INTERNAL_SERVER_ERROR),
CodexErr::ResponseStreamFailed(err) => err.source.status(),
CodexErr::ConnectionFailed(err) => err.source.status(),
_ => None,
}
}
pub fn to_error_event(&self, message_prefix: Option<String>) -> ErrorEvent {
let error_message = self.to_string();
let message: String = match message_prefix {
Some(prefix) => format!("{prefix}: {error_message}"),
None => error_message,
};
ErrorEvent {
message,
http_status_code: http_status_code_value(self.http_status_code()),
}
}
}
pub fn http_status_code_value(http_status_code: Option<StatusCode>) -> Option<u16> {
http_status_code.as_ref().map(StatusCode::as_u16)
}
pub fn get_error_message_ui(e: &CodexErr) -> String {
@@ -775,4 +807,43 @@ mod tests {
assert_eq!(err.to_string(), expected);
});
}
#[test]
fn error_event_includes_http_status_code_when_available() {
let err = CodexErr::UnexpectedStatus(UnexpectedResponseError {
status: StatusCode::BAD_REQUEST,
body: "oops".to_string(),
request_id: Some("req-1".to_string()),
});
let event = err.to_error_event(None);
assert_eq!(
event.message,
"unexpected status 400 Bad Request: oops, request id: req-1"
);
assert_eq!(
event.http_status_code,
Some(StatusCode::BAD_REQUEST.as_u16())
);
}
#[test]
fn error_event_omits_http_status_code_when_unknown() {
let event = CodexErr::Fatal("boom".to_string()).to_error_event(None);
assert_eq!(event.message, "Fatal error: boom");
assert_eq!(event.http_status_code, None);
}
#[test]
fn error_event_applies_message_wrapper() {
let event = CodexErr::Fatal("boom".to_string())
.to_error_event(Some("Error running remote compact task".to_string()));
assert_eq!(
event.message,
"Error running remote compact task: Fatal error: boom"
);
assert_eq!(event.http_status_code, None);
}
}
+1 -1
View File
@@ -72,7 +72,7 @@ For complete documentation of the `Op` and `EventMsg` variants, refer to [protoc
- `EventMsg::AgentMessage` Messages from the `Model`
- `EventMsg::ExecApprovalRequest` Request approval from user to execute a command
- `EventMsg::TaskComplete` A task completed successfully
- `EventMsg::Error` A task stopped with an error
- `EventMsg::Error` A task stopped with an error (includes an optional `http_status_code` when available)
- `EventMsg::Warning` A non-fatal warning that the client should surface to the user
- `EventMsg::TurnComplete` Contains a `response_id` bookmark for last `response_id` executed by the task. This can be used to continue the task at a later point in time, perhaps with additional user input.
@@ -161,7 +161,7 @@ impl EventProcessor for EventProcessorWithHumanOutput {
fn process_event(&mut self, event: Event) -> CodexStatus {
let Event { id: _, msg } = event;
match msg {
EventMsg::Error(ErrorEvent { message }) => {
EventMsg::Error(ErrorEvent { message, .. }) => {
let prefix = "ERROR:".style(self.red);
ts_msg!(self, "{prefix} {message}");
}
@@ -221,7 +221,7 @@ impl EventProcessor for EventProcessorWithHumanOutput {
EventMsg::BackgroundEvent(BackgroundEventEvent { message }) => {
ts_msg!(self, "{}", message.style(self.dimmed));
}
EventMsg::StreamError(StreamErrorEvent { message }) => {
EventMsg::StreamError(StreamErrorEvent { message, .. }) => {
ts_msg!(self, "{}", message.style(self.dimmed));
}
EventMsg::TaskStarted(_) => {
@@ -539,6 +539,7 @@ fn error_event_produces_error() {
"e1",
EventMsg::Error(codex_core::protocol::ErrorEvent {
message: "boom".to_string(),
http_status_code: Some(500),
}),
));
assert_eq!(
@@ -578,6 +579,7 @@ fn stream_error_event_produces_error() {
"e1",
EventMsg::StreamError(codex_core::protocol::StreamErrorEvent {
message: "retrying".to_string(),
http_status_code: Some(500),
}),
));
assert_eq!(
@@ -596,6 +598,7 @@ fn error_followed_by_task_complete_produces_turn_failed() {
"e1",
EventMsg::Error(ErrorEvent {
message: "boom".to_string(),
http_status_code: Some(500),
}),
);
assert_eq!(
+4
View File
@@ -686,6 +686,8 @@ pub struct ExitedReviewModeEvent {
#[derive(Debug, Clone, Deserialize, Serialize, JsonSchema, TS)]
pub struct ErrorEvent {
pub message: String,
#[serde(default)]
pub http_status_code: Option<u16>,
}
#[derive(Debug, Clone, Deserialize, Serialize, JsonSchema, TS)]
@@ -1363,6 +1365,8 @@ pub struct UndoCompletedEvent {
#[derive(Debug, Clone, Deserialize, Serialize, JsonSchema, TS)]
pub struct StreamErrorEvent {
pub message: String,
#[serde(default)]
pub http_status_code: Option<u16>,
}
#[derive(Debug, Clone, Deserialize, Serialize, JsonSchema, TS)]
+4 -2
View File
@@ -1627,7 +1627,7 @@ impl ChatWidget {
self.on_rate_limit_snapshot(ev.rate_limits);
}
EventMsg::Warning(WarningEvent { message }) => self.on_warning(message),
EventMsg::Error(ErrorEvent { message }) => self.on_error(message),
EventMsg::Error(ErrorEvent { message, .. }) => self.on_error(message),
EventMsg::McpStartupUpdate(ev) => self.on_mcp_startup_update(ev),
EventMsg::McpStartupComplete(ev) => self.on_mcp_startup_complete(ev),
EventMsg::TurnAborted(ev) => match ev.reason {
@@ -1670,7 +1670,9 @@ impl ChatWidget {
}
EventMsg::UndoStarted(ev) => self.on_undo_started(ev),
EventMsg::UndoCompleted(ev) => self.on_undo_completed(ev),
EventMsg::StreamError(StreamErrorEvent { message }) => self.on_stream_error(message),
EventMsg::StreamError(StreamErrorEvent { message, .. }) => {
self.on_stream_error(message)
}
EventMsg::UserMessage(ev) => {
if from_replay {
self.on_user_message_event(ev);
+1
View File
@@ -2502,6 +2502,7 @@ fn stream_error_updates_status_indicator() {
id: "sub-1".into(),
msg: EventMsg::StreamError(StreamErrorEvent {
message: msg.to_string(),
http_status_code: None,
}),
});