clone/codex
1
0
Fork 0
mirror of https://github.com/openai/codex.git synced 2026-05-01 09:56:37 +00:00
Code Issues Packages Projects Releases Wiki Activity

analytics: reduce responses api call facts

Browse Source
This commit is contained in:
rhan-oai
2026-04-17 18:31:25 -07:00
parent f42ab32397
commit d783b14963
1 changed files with 165 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

186
codex-rs/analytics/src/analytics_client_tests.rs
View File

@@ -7,6 +7,8 @@ use crate::events::CodexCompactionEventRequest;
use crate::events::CodexHookRunEventRequest;
use crate::events::CodexPluginEventRequest;
use crate::events::CodexPluginUsedEventRequest;
use crate::events::CodexResponsesApiCallEventParams;
use crate::events::CodexResponsesApiCallEventRequest;
use crate::events::CodexRuntimeMetadata;
use crate::events::CodexTurnEventRequest;
use crate::events::ThreadInitializedEvent;
@@ -49,8 +51,13 @@ use crate::facts::TurnTokenUsageFact;
use crate::reducer::AnalyticsReducer;
use crate::reducer::normalize_path_for_skill_id;
use crate::reducer::skill_id_for_local_skill;
use crate::responses_api::CodexResponseItemRole;
use crate::responses_api::CodexResponseItemType;
use crate::responses_api::CodexResponseMessagePhase;
use crate::responses_api::CodexResponsesApiCallFact;
use crate::responses_api::CodexResponsesApiCallStatus;
use crate::responses_api::CodexResponsesApiItemMetadata;
use crate::responses_api::CodexResponsesApiItemPhase;
use codex_app_server_protocol::ApprovalsReviewer as AppServerApprovalsReviewer;
use codex_app_server_protocol::AskForApproval as AppServerAskForApproval;
use codex_app_server_protocol::ClientInfo;
@@ -268,6 +275,30 @@ fn sample_turn_token_usage_fact(thread_id: &str, turn_id: &str) -> TurnTokenUsag
}
}
fn sample_responses_api_item_metadata(
item_phase: CodexResponsesApiItemPhase,
item_index: usize,
) -> CodexResponsesApiItemMetadata {
CodexResponsesApiItemMetadata {
item_phase,
item_index,
response_item_type: CodexResponseItemType::Message,
role: Some(CodexResponseItemRole::Assistant),
status: Some("completed".to_string()),
message_phase: Some(CodexResponseMessagePhase::FinalAnswer),
call_id: None,
tool_name: None,
content_bytes: Some(12),
arguments_bytes: None,
output_bytes: None,
encrypted_content_bytes: None,
image_generation_revised_prompt_bytes: None,
image_generation_result_bytes: None,
text_part_count: Some(1),
image_part_count: Some(0),
}
}
fn sample_turn_completed_notification(
thread_id: &str,
turn_id: &str,
@@ -747,6 +778,102 @@ fn compaction_event_serializes_expected_shape() {
);
}
#[test]
fn responses_api_call_event_serializes_expected_shape() {
let event = TrackEventRequest::ResponsesApiCall(Box::new(CodexResponsesApiCallEventRequest {
event_type: "codex_responses_api_call_event",
event_params: CodexResponsesApiCallEventParams {
thread_id: "thread-1".to_string(),
turn_id: "turn-1".to_string(),
ephemeral: false,
thread_source: Some("user".to_string()),
initialization_mode: ThreadInitializationMode::New,
subagent_source: None,
parent_thread_id: None,
app_server_client: sample_app_server_client_metadata(),
runtime: sample_runtime_metadata(),
responses_id: Some("resp_123".to_string()),
turn_responses_call_index: 2,
model: Some("gpt-5".to_string()),
model_provider: Some("openai".to_string()),
reasoning_effort: Some("high".to_string()),
status: CodexResponsesApiCallStatus::Completed,
error: None,
started_at: 100,
completed_at: Some(102),
duration_ms: Some(2345),
input_item_count: 1,
output_item_count: 1,
input_tokens: Some(10),
cached_input_tokens: Some(4),
output_tokens: Some(20),
reasoning_output_tokens: Some(3),
total_tokens: Some(30),
items: vec![
sample_responses_api_item_metadata(
CodexResponsesApiItemPhase::Input,
/*item_index*/ 0,
),
sample_responses_api_item_metadata(
CodexResponsesApiItemPhase::Output,
/*item_index*/ 0,
),
],
},
}));
let payload = serde_json::to_value(&event).expect("serialize responses api call event");
assert_eq!(payload["event_type"], "codex_responses_api_call_event");
assert_eq!(payload["event_params"]["thread_id"], "thread-1");
assert_eq!(payload["event_params"]["turn_id"], "turn-1");
assert_eq!(payload["event_params"]["ephemeral"], false);
assert_eq!(payload["event_params"]["thread_source"], "user");
assert_eq!(payload["event_params"]["initialization_mode"], "new");
assert_eq!(
payload["event_params"]["app_server_client"]["product_client_id"],
DEFAULT_ORIGINATOR
);
assert_eq!(
payload["event_params"]["app_server_client"]["rpc_transport"],
"stdio"
);
assert_eq!(
payload["event_params"]["runtime"]["codex_rs_version"],
"0.1.0"
);
assert_eq!(payload["event_params"]["responses_id"], "resp_123");
assert_eq!(payload["event_params"]["turn_responses_call_index"], 2);
assert_eq!(payload["event_params"]["model"], "gpt-5");
assert_eq!(payload["event_params"]["model_provider"], "openai");
assert_eq!(payload["event_params"]["reasoning_effort"], "high");
assert_eq!(payload["event_params"]["status"], "completed");
assert_eq!(payload["event_params"]["error"], json!(null));
assert_eq!(payload["event_params"]["started_at"], 100);
assert_eq!(payload["event_params"]["completed_at"], 102);
assert_eq!(payload["event_params"]["duration_ms"], 2345);
assert_eq!(payload["event_params"]["input_item_count"], 1);
assert_eq!(payload["event_params"]["output_item_count"], 1);
assert_eq!(payload["event_params"]["input_tokens"], 10);
assert_eq!(payload["event_params"]["cached_input_tokens"], 4);
assert_eq!(payload["event_params"]["output_tokens"], 20);
assert_eq!(payload["event_params"]["reasoning_output_tokens"], 3);
assert_eq!(payload["event_params"]["total_tokens"], 30);
assert_eq!(payload["event_params"]["items"][0]["item_phase"], "input");
assert_eq!(
payload["event_params"]["items"][0]["response_item_type"],
"message"
);
assert_eq!(payload["event_params"]["items"][0]["role"], "assistant");
assert_eq!(
payload["event_params"]["items"][0]["message_phase"],
"final_answer"
);
assert_eq!(payload["event_params"]["items"][0]["text_part_count"], 1);
assert_eq!(payload["event_params"]["items"][0]["image_part_count"], 0);
assert_eq!(payload["event_params"]["items"][1]["item_phase"], "output");
}
#[test]
fn app_used_dedupe_is_keyed_by_turn_and_connector() {
let (sender, _receiver) = mpsc::channel(1);
@@ -1056,6 +1183,7 @@ async fn compaction_event_ingests_custom_fact() {
async fn responses_api_call_event_ingests_custom_fact() {
let mut reducer = AnalyticsReducer::default();
let mut events = Vec::new();
ingest_turn_prerequisites(
&mut reducer,
&mut events,
@@ -1065,7 +1193,7 @@ async fn responses_api_call_event_ingests_custom_fact() {
/*include_token_usage*/ false,
)
.await;
assert!(events.is_empty());
events.clear();
reducer
.ingest(
@@ -1073,21 +1201,30 @@ async fn responses_api_call_event_ingests_custom_fact() {
CodexResponsesApiCallFact {
thread_id: "thread-2".to_string(),
turn_id: "turn-2".to_string(),
responses_id: Some("resp-1".to_string()),
turn_responses_call_index: 2,
responses_id: Some("resp_456".to_string()),
turn_responses_call_index: 1,
status: CodexResponsesApiCallStatus::Completed,
error: None,
started_at: 100,
completed_at: Some(101),
duration_ms: Some(1200),
input_item_count: 3,
output_item_count: 4,
input_tokens: Some(10),
cached_input_tokens: Some(2),
output_tokens: Some(6),
reasoning_output_tokens: Some(1),
total_tokens: Some(16),
items: Vec::new(),
input_item_count: 1,
output_item_count: 1,
input_tokens: Some(123),
cached_input_tokens: Some(45),
output_tokens: Some(140),
reasoning_output_tokens: Some(13),
total_tokens: Some(321),
items: vec![
sample_responses_api_item_metadata(
CodexResponsesApiItemPhase::Input,
/*item_index*/ 0,
),
sample_responses_api_item_metadata(
CodexResponsesApiItemPhase::Output,
/*item_index*/ 0,
),
],
},
))),
&mut events,
@@ -1099,22 +1236,29 @@ async fn responses_api_call_event_ingests_custom_fact() {
assert_eq!(payload[0]["event_type"], "codex_responses_api_call_event");
assert_eq!(payload[0]["event_params"]["thread_id"], "thread-2");
assert_eq!(payload[0]["event_params"]["turn_id"], "turn-2");
assert_eq!(payload[0]["event_params"]["responses_id"], "resp-1");
assert_eq!(payload[0]["event_params"]["turn_responses_call_index"], 2);
assert_eq!(payload[0]["event_params"]["model"], "gpt-5");
assert_eq!(payload[0]["event_params"]["model_provider"], "openai");
assert_eq!(payload[0]["event_params"]["status"], "completed");
assert_eq!(payload[0]["event_params"]["input_tokens"], 10);
assert_eq!(payload[0]["event_params"]["cached_input_tokens"], 2);
assert_eq!(payload[0]["event_params"]["output_tokens"], 6);
assert_eq!(payload[0]["event_params"]["reasoning_output_tokens"], 1);
assert_eq!(payload[0]["event_params"]["total_tokens"], 16);
assert_eq!(payload[0]["event_params"]["ephemeral"], false);
assert_eq!(payload[0]["event_params"]["thread_source"], "user");
assert_eq!(payload[0]["event_params"]["initialization_mode"], "new");
assert_eq!(
payload[0]["event_params"]["app_server_client"]["product_client_id"],
"codex-tui"
);
assert_eq!(payload[0]["event_params"]["runtime"]["runtime_os"], "macos");
assert_eq!(payload[0]["event_params"]["responses_id"], "resp_456");
assert_eq!(payload[0]["event_params"]["turn_responses_call_index"], 1);
assert_eq!(payload[0]["event_params"]["model"], "gpt-5");
assert_eq!(payload[0]["event_params"]["model_provider"], "openai");
assert_eq!(payload[0]["event_params"]["reasoning_effort"], json!(null));
assert_eq!(payload[0]["event_params"]["status"], "completed");
assert_eq!(payload[0]["event_params"]["input_tokens"], 123);
assert_eq!(
payload[0]["event_params"]["items"][0]["item_phase"],
"input"
);
assert_eq!(
payload[0]["event_params"]["items"][1]["item_phase"],
"output"
);
}
#[test]