We added retrieval event logging and lineage-oriented artifacts. The goal is operational clarity: a run should leave enough trace data to understand what happened.
This is not user-facing analytics. It is engineering evidence for debugging retrieval decisions, cache behavior, and grounding failures.
We clarified the boundary between conversation memory, retrieved evidence, and answer synthesis in the Agentic RAG runtime.
We tightened how retrieval diagnostics, evidence boundaries, and answer synthesis traces connect in the Agentic RAG loop.
Datasource routing work was added so retrieval can target relevant sources instead of searching every corpus the same way.