GjSrSSKJr SSKJrJrJr SSKJr SSK J r \(aSSK J r \"S\RS9r"S S 5rg) a Experimental client-side task support. This module provides client methods for interacting with MCP tasks. WARNING: These APIs are experimental and may change without notice. Example: # Call a tool as a task result = await session.experimental.call_tool_as_task("tool_name", {"arg": "value"}) task_id = result.task.taskId # Get task status status = await session.experimental.get_task(task_id) # Get task result when complete if status.status == "completed": result = await session.experimental.get_task_result(task_id, CallToolResult) # List all tasks tasks = await session.experimental.list_tasks() # Cancel a task await session.experimental.cancel_task(task_id) ) AsyncIterator) TYPE_CHECKINGAnyTypeVarN)poll_until_terminal) ClientSessionResultT)boundc J\rSrSrSrSSjrSSSS.S\S \\\4S-S \ S \\\4S-S\ R4 S jjjr S \S\ R4SjrS \S\\S\4SjrSS\S-S\ R$4SjjrS \S\ R(4SjrS \S\\ R4SjrSrg)ExperimentalClientFeatures'z Experimental client features for tasks and other experimental APIs. WARNING: These APIs are experimental and may change without notice. Access via session.experimental: status = await session.experimental.get_task(task_id) returnNcXlgN_session)selfsessions _/Users/admin/workspace/tools/venv/lib/python3.13/site-packages/mcp/client/experimental/tasks.py__init__#ExperimentalClientFeatures.__init__1s i`)ttlmetaname argumentsrrcV# SnUb [RR"S0UD6nURR [R "[R "[R"UU[R"US9US9S95[R5IShvN $N7f)aCall a tool as a task, returning a CreateTaskResult for polling. This is a convenience method for calling tools that support task execution. The server will return a task reference instead of the immediate result, which can then be polled via `get_task()` and retrieved via `get_task_result()`. Args: name: The tool name arguments: Tool arguments ttl: Task time-to-live in milliseconds (default: 60000 = 1 minute) meta: Optional metadata to include in the request Returns: CreateTaskResult containing the task reference Example: # Create task result = await session.experimental.call_tool_as_task( "long_running_tool", {"input": "data"} ) task_id = result.task.taskId # Poll for completion while True: status = await session.experimental.get_task(task_id) if status.status == "completed": break await asyncio.sleep(0.5) # Get result final = await session.experimental.get_task_result(task_id, CallToolResult) N)r)rrtask_metaparams) types RequestParamsMetar send_request ClientRequestCallToolRequestCallToolRequestParams TaskMetadataCreateTaskResult)rrrrrrs rcall_tool_as_task,ExperimentalClientFeatures.call_tool_as_task4sP26  '',,4t4E]]//   %% 66!"+"//C8#    " "    sB B)"B'#B)task_idc # URR[R"[R"[R "US9S95[R 5IShvN $N7f)z Get the current status of a task. Args: task_id: The task identifier Returns: GetTaskResult containing the task status and metadata taskIdr N)rr&r#r'GetTaskRequestGetTaskRequestParams GetTaskResultrr.s rget_task#ExperimentalClientFeatures.get_taskns]]]//   $$ 55WE         A&A/(A-)A/ result_typec # URR[R"[R"[R "US9S95U5IShvN $N7f)ah Get the result of a completed task. The result type depends on the original request type: - tools/call tasks return CallToolResult - Other request types return their corresponding result type Args: task_id: The task identifier result_type: The expected result type (e.g., CallToolResult) Returns: The task result, validated against result_type r0r N)rr&r#r'GetTaskPayloadRequestGetTaskPayloadRequestParams)rr.r9s rget_task_result*ExperimentalClientFeatures.get_task_resultsW&]]//   ++ <D(8  8 38nt# 8    8 t c e.A.A &  ']    <" d      *  1G1G &s}U=P=P/Qrr )rTcollections.abcrtypingrrr mcp.typesr#%mcp.shared.experimental.tasks.pollingrmcp.client.sessionrResultr r r"rrr`s=4*..E0 )5<< 0yyr