Async tool support

Are async functions currently unsupported, or is there a workaround for tools that rely on asynchronous behaviour?

For example, I have a tool that retrieves data asynchronously from various web resources, and the agent needs to await these responses:

@tool
async def retrieve_datasets(query: str, retriever: RepoRetriever) -> dict:
    """
    This tool can be used to retrieve metadata from different repositories.

    Note this is a async function and needs to be awaited.

    Args:
        query: The query string to search for.
        retriever: An instance of the RepoRetriever class.
    """
    response = await(retriever.query_all_repos(query))

    top_5_results = response.get("results", [])[:5]
    return top_5_results

Attempting to use this tool results in the following error:

─ Executing this code: ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
 precipitation_data = await retrieve_datasets(query="precipitation data", retriever=retriever)
 print(precipitation_data)
───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
Code execution failed at line 'precipitation_data = await retrieve_datasets(query="precipitation data", retriever=retriever)' because of the following error:
Await is not supported.

Is there a recommended approach to handle asynchronous tools in such scenarios? Any guidance would be greatly appreciated.

Start by reproducing the example async tool and tracing the tool execution path to the code-execution handling that reports "Await is not supported." Check existing tool-execution tests or documentation for related behavior; done should mean async tools have a defined supported behavior or a documented workaround, backed by a regression test if execution behavior changes.