fix: stream stage events live, thread source_chunks, fix LangChain wrapper

fix: mypy errors in agent factory (model_name, type annotation)

fix: deferred imports, match iteration budget, token cost tracking

feat: langchain evaluation CLI script and Makefile target

feat: langchain evaluation runner producing EvalResult objects

feat: langchain tool-calling agent factory

fix: ruff violations and sync fallback safety for nested event loops

feat: langchain search tool with metadata capture + calculator

feat: langchain retriever wrapper over existing async hybrid retriever