Skip to content

The fastest, cheapest, most efficient open-source agent harness. Run more than 500 agents on a $10 VPS.

feat(inference): expose summarizer model id on turn models

Steven Enamakelab047b0ae71e
diff --git a/crates/openhuman-core/src/inference/turn_models.rs b/crates/openhuman-core/src/inference/turn_models.rs
index 026ba4f8fb..9b8dd11957 100644
--- a/crates/openhuman-core/src/inference/turn_models.rs
+++ b/crates/openhuman-core/src/inference/turn_models.rs
@@ -20,6 +20,9 @@ pub(crate) struct TurnModels {
/// A model for the context-window summarizer (a distinct adapter instance so
/// its provider errors don't touch the turn's `error_slot`).
pub(crate) summarizer: Arc<dyn tinyinference_llm::model::ChatModel<()>>,
+ /// The model id [`summarizer`](Self::summarizer) calls, as the compaction
+ /// summarizer reports it in its logs and summary provenance.
+ pub(crate) summarizer_model_id: String,
/// Recovers the primary's original (downcastable) provider error on failure.
pub(crate) error_slot: tinyinference_llm::model::ModelErrorSlot,
/// Native owner set retained so its route and primary error slots are
@@ -77,6 +80,11 @@ impl TurnModels {
self
}
+ /// The model id the context-compaction summarizer calls.
+ pub(crate) fn summarizer_model_id(&self) -> &str {
+ &self.summarizer_model_id
+ }
+
/// Provider telemetry id for this turn (`{provider_id}.{model}`).
pub(crate) fn provider_id(&self) -> &str {
&self.provider_id
@@ -197,6 +205,7 @@ fn build_turn_models_crate(
primary,
routes,
summarizer,
+ summarizer_model_id: model.to_owned(),
error_slot,
native_model_set: Some(built),
provider_id,
@@ -466,6 +475,7 @@ impl TurnModelSource {
primary,
routes: Vec::new(),
summarizer,
+ summarizer_model_id: model.to_owned(),
error_slot: Arc::new(std::sync::Mutex::new(None)),
native_model_set: None,
provider_id,