modelcontextprotocol · maxisbey · Mar 24, 2026 · Mar 22, 2026 · Mar 24, 2026
diff --git a/src/mcp/server/lowlevel/server.py b/src/mcp/server/lowlevel/server.py
@@ -671,16 +671,23 @@ async def run(
                 await stack.enter_async_context(task_support.run())
 
             async with anyio.create_task_group() as tg:
-                async for message in session.incoming_messages:
-                    logger.debug("Received message: %s", message)
-
-                    tg.start_soon(
-                        self._handle_message,
-                        message,
-                        session,
-                        lifespan_context,
-                        raise_exceptions,
-                    )
+                try:
+                    async for message in session.incoming_messages:
+                        logger.debug("Received message: %s", message)
+
+                        tg.start_soon(
+                            self._handle_message,
+                            message,
+                            session,
+                            lifespan_context,
+                            raise_exceptions,
+                        )
+                finally:
+                    # Transport closed: cancel in-flight handlers. Without this the
+                    # TG join waits for them, and when they eventually try to
+                    # respond they hit a closed write stream (the session's
+                    # _receive_loop closed it when the read stream ended).
+                    tg.cancel_scope.cancel()
 
     async def _handle_message(
         self,
@@ -763,12 +770,18 @@ async def _handle_request(
                 response = await handler(req)
             except McpError as err:  # pragma: no cover
                 response = err.error
-            except anyio.get_cancelled_exc_class():  # pragma: no cover
-                logger.info(
-                    "Request %s cancelled - duplicate response suppressed",
-                    message.request_id,
-                )
-                return
+            except anyio.get_cancelled_exc_class():
+                if message.cancelled:
+                    # Client sent CancelledNotification; responder.cancel() already
+                    # sent an error response, so skip the duplicate.
+                    logger.info(
+                        "Request %s cancelled - duplicate response suppressed",
+                        message.request_id,
+                    )
+                    return
+                # Transport-close cancellation from the TG in run(); re-raise so the
+                # TG swallows its own cancellation.
+                raise
             except Exception as err:  # pragma: no cover
                 if raise_exceptions:
                     raise err
@@ -777,16 +790,24 @@ async def _handle_request(
                 # Reset the global state after we are done
                 if token is not None:  # pragma: no branch
                     request_ctx.reset(token)
-
-            await message.respond(response)
         else:  # pragma: no cover
-            await message.respond(
-                types.ErrorData(
-                    code=types.METHOD_NOT_FOUND,
-                    message="Method not found",
-                )
+            response = types.ErrorData(
+                code=types.METHOD_NOT_FOUND,
+                message="Method not found",
             )
 
+        try:
+            await message.respond(response)
+        except (anyio.BrokenResourceError, anyio.ClosedResourceError):
+            # Transport closed between handler unblocking and respond. Happens
+            # when _receive_loop's finally wakes a handler blocked on
+            # send_request: the handler runs to respond() before run()'s TG
+            # cancel fires, but after the write stream closed. Closed if our
+            # end closed (_receive_loop's async-with exit); Broken if the peer
+            # end closed first (streamable_http terminate()).
+            logger.debug("Response for %s dropped - transport closed", message.request_id)
+            return
+
         logger.debug("Response sent")
 
     async def _handle_notification(self, notify: Any):

diff --git a/src/mcp/shared/session.py b/src/mcp/shared/session.py
@@ -108,7 +108,7 @@ def __exit__(
     ) -> None:
         """Exit the context manager, performing cleanup and notifying completion."""
         try:
-            if self._completed:  # pragma: no branch
+            if self._completed:
                 self._on_complete(self)
         finally:
             self._entered = False
@@ -445,7 +445,9 @@ async def _receive_loop(self) -> None:
             finally:
                 # after the read stream is closed, we need to send errors
                 # to any pending requests
-                for id, stream in self._response_streams.items():
+                # Snapshot: stream.send() wakes the waiter, whose finally pops
+                # from _response_streams before the next __next__() call.
+                for id, stream in list(self._response_streams.items()):
                     error = ErrorData(code=CONNECTION_CLOSED, message="Connection closed")
                     try:
                         await stream.send(JSONRPCError(jsonrpc="2.0", id=id, error=error))

diff --git a/tests/server/test_cancel_handling.py b/tests/server/test_cancel_handling.py
@@ -9,14 +9,21 @@
 from mcp.server.lowlevel.server import Server
 from mcp.shared.exceptions import McpError
 from mcp.shared.memory import create_connected_server_and_client_session
+from mcp.shared.message import SessionMessage
 from mcp.types import (
+    LATEST_PROTOCOL_VERSION,
     CallToolRequest,
     CallToolRequestParams,
     CallToolResult,
     CancelledNotification,
     CancelledNotificationParams,
+    ClientCapabilities,
     ClientNotification,
     ClientRequest,
+    Implementation,
+    InitializeRequestParams,
+    JSONRPCNotification,
+    JSONRPCRequest,
     Tool,
 )
 
@@ -108,3 +115,156 @@ async def first_request():
         assert isinstance(content, types.TextContent)
         assert content.text == "Call number: 2"
         assert call_count == 2
+
+
+@pytest.mark.anyio
+async def test_server_cancels_in_flight_handlers_on_transport_close():
+    """When the transport closes mid-request, server.run() must cancel in-flight
+    handlers rather than join on them.
+
+    Without the cancel, the task group waits for the handler, which then tries
+    to respond through a write stream that _receive_loop already closed,
+    raising ClosedResourceError and crashing server.run() with exit code 1.
+
+    This drives server.run() with raw memory streams because InMemoryTransport
+    wraps it in its own finally-cancel (_memory.py) which masks the bug.
+    """
+    handler_started = anyio.Event()
+    handler_cancelled = anyio.Event()
+    server_run_returned = anyio.Event()
+
+    server = Server("test")
+
+    @server.call_tool()
+    async def handle_call_tool(name: str, arguments: dict[str, Any] | None) -> list[types.TextContent]:
+        handler_started.set()
+        try:
+            await anyio.sleep_forever()
+        finally:
+            handler_cancelled.set()
+        # unreachable: sleep_forever only exits via cancellation
+        raise AssertionError  # pragma: no cover
+
+    to_server, server_read = anyio.create_memory_object_stream[SessionMessage | Exception](10)
+    server_write, from_server = anyio.create_memory_object_stream[SessionMessage](10)
+
+    async def run_server():
+        await server.run(server_read, server_write, server.create_initialization_options())
+        server_run_returned.set()
+
+    init_req = JSONRPCRequest(
+        jsonrpc="2.0",
+        id=1,
+        method="initialize",
+        params=InitializeRequestParams(
+            protocolVersion=LATEST_PROTOCOL_VERSION,
+            capabilities=ClientCapabilities(),
+            clientInfo=Implementation(name="test", version="1.0"),
+        ).model_dump(by_alias=True, mode="json", exclude_none=True),
+    )
+    initialized = JSONRPCNotification(jsonrpc="2.0", method="notifications/initialized")
+    call_req = JSONRPCRequest(
+        jsonrpc="2.0",
+        id=2,
+        method="tools/call",
+        params=CallToolRequestParams(name="slow", arguments={}).model_dump(by_alias=True, mode="json"),
+    )
+
+    with anyio.fail_after(5):
+        async with anyio.create_task_group() as tg, to_server, server_read, server_write, from_server:
+            tg.start_soon(run_server)
+
+            await to_server.send(SessionMessage(message=types.JSONRPCMessage(init_req)))
+            await from_server.receive()  # init response
+            await to_server.send(SessionMessage(message=types.JSONRPCMessage(initialized)))
+            await to_server.send(SessionMessage(message=types.JSONRPCMessage(call_req)))
+
+            await handler_started.wait()
+
+            # Close the server's input stream — this is what stdin EOF does.
+            # server.run()'s incoming_messages loop ends, finally-cancel fires,
+            # handler gets CancelledError, server.run() returns.
+            await to_server.aclose()
+
+            await server_run_returned.wait()
+
+    assert handler_cancelled.is_set()
+
+
+@pytest.mark.anyio
+async def test_server_handles_transport_close_with_pending_server_to_client_requests():
+    """When the transport closes while handlers are blocked on server→client
+    requests (sampling, roots, elicitation), server.run() must still exit cleanly.
+
+    Two bugs covered:
+      1. _receive_loop's finally iterates _response_streams with await checkpoints
+         inside; the woken handler's send_request finally pops from that dict
+         before the next __next__() — RuntimeError: dictionary changed size.
+      2. The woken handler's MCPError is caught in _handle_request, which falls
+         through to respond() against a write stream _receive_loop already closed.
+    """
+    handlers_started = 0
+    both_started = anyio.Event()
+    server_run_returned = anyio.Event()
+
+    server = Server("test")
+
+    @server.call_tool()
+    async def handle_call_tool(name: str, arguments: dict[str, Any] | None) -> list[types.TextContent]:
+        nonlocal handlers_started
+        handlers_started += 1
+        if handlers_started == 2:
+            both_started.set()
+        # Blocks on send_request waiting for a client response that never comes.
+        # _receive_loop's finally will wake this with CONNECTION_CLOSED.
+        await server.request_context.session.list_roots()
+        raise AssertionError  # pragma: no cover
+
+    to_server, server_read = anyio.create_memory_object_stream[SessionMessage | Exception](10)
+    server_write, from_server = anyio.create_memory_object_stream[SessionMessage](10)
+
+    async def run_server():
+        await server.run(server_read, server_write, server.create_initialization_options())
+        server_run_returned.set()
+
+    init_req = JSONRPCRequest(
+        jsonrpc="2.0",
+        id=1,
+        method="initialize",
+        params=InitializeRequestParams(
+            protocolVersion=LATEST_PROTOCOL_VERSION,
+            capabilities=ClientCapabilities(),
+            clientInfo=Implementation(name="test", version="1.0"),
+        ).model_dump(by_alias=True, mode="json", exclude_none=True),
+    )
+    initialized = JSONRPCNotification(jsonrpc="2.0", method="notifications/initialized")
+
+    with anyio.fail_after(5):
+        async with anyio.create_task_group() as tg, to_server, server_read, server_write, from_server:
+            tg.start_soon(run_server)
+
+            await to_server.send(SessionMessage(message=types.JSONRPCMessage(init_req)))
+            await from_server.receive()  # init response
+            await to_server.send(SessionMessage(message=types.JSONRPCMessage(initialized)))
+
+            # Two tool calls → two handlers → two _response_streams entries.
+            for rid in (2, 3):
+                call_req = JSONRPCRequest(
+                    jsonrpc="2.0",
+                    id=rid,
+                    method="tools/call",
+                    params=CallToolRequestParams(name="t", arguments={}).model_dump(by_alias=True, mode="json"),
+                )
+                await to_server.send(SessionMessage(message=types.JSONRPCMessage(call_req)))
+
+            await both_started.wait()
+            # Drain the two roots/list requests so send_request's _write_stream.send()
+            # completes and both handlers are parked at response_stream_reader.receive().
+            await from_server.receive()
+            await from_server.receive()
+
+            await to_server.aclose()
+
+            # Without the fixes: RuntimeError (dict mutation) or ClosedResourceError
+            # (respond after write-stream close) escapes run_server and this hangs.
+            await server_run_returned.wait()