ai-agents-vector-store-async-opensearch-vector-store-operations-block-the-event-c0cc11c5

THEN Create async subclasses of OpensearchVectorClient and OpensearchVectorStore that override query methods with async/await and use the AsyncOpenSearch client from the opensearch-py library. Monkeypatch the internal client factory to return an async client.

class AsyncOpensearchVectorClient(OpensearchVectorClient):
    async def aquery(self, query: VectorStoreQuery) -> VectorStoreQueryResult:
        # identical to sync method but with async/await on opensearch calls
        ...
# Monkeypatch _get_opensearch_client to return AsyncOpenSearch(client_url, **kwargs)