Class●Since v0.4

AsyncSubAgentMiddleware

Middleware for async subagents running on remote LangGraph servers.

This middleware adds tools for launching, monitoring, and updating background jobs on remote LangGraph deployments. Unlike the synchronous SubAgentMiddleware, async subagents return immediately with a job ID, allowing the main agent to continue working while subagents execute.

Job IDs are persisted in the agent state under async_subagent_jobs so they survive context compaction/offloading and can be accessed programmatically.

AsyncSubAgentMiddleware(
  self,
  *,
  async_subagents: list[AsyncSubAgent],
  system_prompt: str | None = ASYNC_TASK_SYSTEM_PROMPT
)

Bases

AgentMiddleware[Any, ContextT, ResponseT]

Example:

from deepagents.middleware.async_subagents import AsyncSubAgentMiddleware

middleware = AsyncSubAgentMiddleware(
    async_subagents=[
        {
            "name": "researcher",
            "description": "Research agent for deep analysis",
            "url": "https://my-deployment.langsmith.dev",
            "graph_id": "research_agent",
        }
    ],
)

Parameters

Name	Type	Description
`async_subagents`*	`list[AsyncSubAgent]`	List of async subagent specifications. Each must include `name`, `description`, and `graph_id`. `url` is optional — omit it to use ASGI transport for local LangGraph servers.
`system_prompt`	`str \| None`	Default:`ASYNC_TASK_SYSTEM_PROMPT` Instructions appended to the main agent's system prompt about how to use the async subagent tools.