langchain-ai/deepagents · error

Failed to get run status: {e}

Error message

Failed to get run status: {e}

What it means

check_async_task wraps client creation and the runs.get status call in a broad except and returns this message string instead of raising, so the calling agent sees a friendly failure. Any error from get_sync (e.g. missing url) or from the LangGraph SDK while fetching run status surfaces as this text.

Source

Thrown at libs/deepagents/deepagents/middleware/async_subagents.py:423

def _build_check_tool(  # noqa: C901  # complexity from necessary error handling
    clients: _ClientCache,
) -> StructuredTool:
    """Build the `check_async_task` tool."""

    def check_async_task(
        task_id: str,
        runtime: ToolRuntime,
    ) -> str | Command:
        task = _resolve_tracked_task(task_id, runtime)
        if isinstance(task, str):
            return task

        try:
            client = clients.get_sync(task["agent_name"])
            run = client.runs.get(thread_id=task["thread_id"], run_id=task["run_id"])
        except Exception as e:  # noqa: BLE001  # get_sync() may raise ValueError; SDK raises untyped errors
            return f"Failed to get run status: {e}"

        thread_values: dict[str, Any] = {}
        if run["status"] == "success":
            try:
                thread = client.threads.get(thread_id=task["thread_id"])
                thread_values = thread.get("values") or {}
            except Exception as e:  # noqa: BLE001  # LangGraph SDK raises untyped errors
                logger.warning("Failed to fetch thread values for task %s: %s", task["task_id"], e)

        result = _build_check_result(run, task["thread_id"], thread_values)
        return _build_check_command(result, task, runtime.tool_call_id)

    async def acheck_async_task(
        task_id: str,
        runtime: ToolRuntime,
    ) -> str | Command:
        task = _resolve_tracked_task(task_id, runtime)

View on GitHub (pinned to a1af029e6e)

Solutions

1. Check the subagent 'url' is set and the server is reachable
2. Read the embedded exception in the returned message and fix the underlying cause (auth, network, IDs)
3. Use async invocation paths that support ASGI transport when subagents have no url

if subagent_spec.get("url") is None:
    # avoid sync status path for ASGI subagents
    use_async_status_check = True

try:
    client = clients.get_sync(task["agent_name"])
    run = client.runs.get(thread_id=task["thread_id"], run_id=task["run_id"])
except Exception as e:
    logger.warning("run status fetch failed: %s", e)
    return None

When it happens

Trigger: Polling an async task whose subagent has no url (get_sync ValueError); network/auth failure or invalid thread_id/run_id when calling client.runs.get; SDK raising untyped errors.

Common situations: Subagent server restarted or unreachable mid-run; task record references a thread/run that was deleted; spec misconfiguration (url=None) discovered lazily at poll time.

Related errors

• Failed to cancel run: {e}
• Failed while waiting for Vercel sandbox startup.
• StateBackend must be used inside a LangGraph graph execution
• StateBackend requires CONFIG_KEY_READ / CONFIG_KEY_SEND in t
• Async subagent '{name}' has no url configured. ASGI transpor

AI-assisted analysis of langchain-ai/deepagents@a1af029e6e (2026-08-29). Data as JSON: /api/errors/f3a2f3290b5d336f. Report an issue: GitHub.