[codex-analytics] centralize thread analytics state (#20300)

## Why

Several analytics event families need the same per-thread attribution
state: the app-server client/runtime associated with a thread and, for
lifecycle-oriented events, the thread metadata captured during
initialization. Keeping connection ids and lifecycle metadata in
separate maps made each consumer rebuild the same thread context and
made subagent attribution harder to resolve consistently.

## What changed

- Replaces the separate thread connection and metadata maps with one
reducer-owned `threads` map.
- Routes guardian, compaction, turn-steer, and turn analytics through
shared thread-state lookups while preserving turn-origin attribution for
turn events and request-origin attribution for steer events.
- Lets newly observed spawned subagent threads inherit their parent
thread connection so later thread-scoped analytics can resolve through
the same state model.
- Adds regression coverage for standalone `SubAgentThreadStarted`
publication plus the `SubAgentSource::ThreadSpawn` parent fallback
through a thread-scoped consumer that depends on inherited connection
state.

## Verification

- `cargo test -p codex-analytics`

---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed
with [ReviewStack](https://reviewstack.dev/openai/codex/pull/20300).
* #18748
* #18747
* #17090
* #17089
* #20239
* #20515
* #20514
* __->__ #20300
This commit is contained in:
rhan-oai authored and GitHub committed 2026-04-30 18:58:50 -07:00
1 parent 972b819213
commit 6b1b227804
2 files changed
+323 -81

No files matched your search

@@ -303,10 +303,10 @@ fn sample_turn_completed_notification(
})
}
fn sample_turn_resolved_config(turn_id: &str) -> TurnResolvedConfigFact {
fn sample_turn_resolved_config(thread_id: &str, turn_id: &str) -> TurnResolvedConfigFact {
TurnResolvedConfigFact {
turn_id: turn_id.to_string(),
thread_id: "thread-2".to_string(),
thread_id: thread_id.to_string(),
num_input_images: 1,
submission_type: None,
ephemeral: false,
@@ -419,6 +419,38 @@ async fn ingest_rejected_turn_steer(
/*include_started*/ false, /*include_token_usage*/ false,
)
.await;
reducer
.ingest(
AnalyticsFact::Initialize {
connection_id: 8,
params: InitializeParams {
client_info: ClientInfo {
name: "codex-web".to_string(),
title: None,
version: "1.0.0".to_string(),
},
capabilities: None,
},
product_client_id: "codex-web".to_string(),
runtime: sample_runtime_metadata(),
rpc_transport: AppServerRpcTransport::Stdio,
},
out,
)
.await;
reducer
.ingest(
AnalyticsFact::ClientResponse {
connection_id: 8,
request_id: RequestId::Integer(6),
response: Box::new(sample_thread_resume_response(
"thread-2", /*ephemeral*/ false, "gpt-5",
)),
},
out,
)
.await;
out.clear();
reducer
.ingest(
AnalyticsFact::ClientRequest {
@@ -519,7 +551,7 @@ async fn ingest_turn_prerequisites(
reducer
.ingest(
AnalyticsFact::Custom(CustomAnalyticsFact::TurnResolvedConfig(Box::new(
sample_turn_resolved_config("turn-2"),
sample_turn_resolved_config("thread-2", "turn-2"),
))),
out,
)
@@ -1436,6 +1468,110 @@ async fn subagent_thread_started_publishes_without_initialize() {
assert_eq!(payload[0]["event_params"]["subagent_source"], "review");
}
#[tokio::test]
async fn subagent_thread_started_inherits_parent_connection_for_new_thread() {
let mut reducer = AnalyticsReducer::default();
let mut events = Vec::new();
let parent_thread_id =
codex_protocol::ThreadId::from_string("44444444-4444-4444-4444-444444444444")
.expect("valid parent thread id");
let parent_thread_id_string = parent_thread_id.to_string();
reducer
.ingest(
AnalyticsFact::Initialize {
connection_id: 7,
params: InitializeParams {
client_info: ClientInfo {
name: "parent-client".to_string(),
title: None,
version: "1.0.0".to_string(),
},
capabilities: None,
},
product_client_id: "parent-client".to_string(),
runtime: sample_runtime_metadata(),
rpc_transport: AppServerRpcTransport::Stdio,
},
&mut events,
)
.await;
reducer
.ingest(
AnalyticsFact::ClientResponse {
connection_id: 7,
request_id: RequestId::Integer(1),
response: Box::new(sample_thread_start_response(
&parent_thread_id_string,
/*ephemeral*/ false,
"gpt-5",
)),
},
&mut events,
)
.await;
reducer
.ingest(
AnalyticsFact::Custom(CustomAnalyticsFact::SubAgentThreadStarted(
SubAgentThreadStartedInput {
thread_id: "thread-review".to_string(),
parent_thread_id: None,
product_client_id: "parent-client".to_string(),
client_name: "parent-client".to_string(),
client_version: "1.0.0".to_string(),
model: "gpt-5".to_string(),
ephemeral: false,
subagent_source: SubAgentSource::ThreadSpawn {
parent_thread_id,
depth: 1,
agent_path: None,
agent_nickname: None,
agent_role: None,
},
created_at: 130,
},
)),
&mut events,
)
.await;
events.clear();
reducer
.ingest(
AnalyticsFact::Custom(CustomAnalyticsFact::Compaction(Box::new(
CodexCompactionEvent {
thread_id: "thread-review".to_string(),
turn_id: "turn-compact".to_string(),
trigger: CompactionTrigger::Manual,
reason: CompactionReason::UserRequested,
implementation: CompactionImplementation::Responses,
phase: CompactionPhase::StandaloneTurn,
strategy: CompactionStrategy::Memento,
status: CompactionStatus::Completed,
error: None,
active_context_tokens_before: 131_000,
active_context_tokens_after: 64_000,
started_at: 100,
completed_at: 101,
duration_ms: Some(1200),
},
))),
&mut events,
)
.await;
let payload = serde_json::to_value(&events).expect("serialize events");
assert_eq!(
payload[0]["event_params"]["app_server_client"]["product_client_id"],
"parent-client"
);
assert_eq!(
payload[0]["event_params"]["parent_thread_id"],
"44444444-4444-4444-4444-444444444444"
);
}
#[test]
fn plugin_used_event_serializes_expected_shape() {
let tracking = TrackEventsContext {
@@ -2149,7 +2285,7 @@ async fn turn_start_error_response_discards_pending_start_request() {
reducer
.ingest(
AnalyticsFact::Custom(CustomAnalyticsFact::TurnResolvedConfig(Box::new(
sample_turn_resolved_config("turn-2"),
sample_turn_resolved_config("thread-2", "turn-2"),
))),
&mut out,
)