Files
Max Isbey ec5b2258c9 Cut import and startup cost with deferred model builds and lazy imports
Every import path now pays only for what it uses, with no public API
added or removed:

- The protocol models (mcp.types / mcp_types, incl. the JSON-RPC
  envelopes and the generated per-version wire packages) build their
  pydantic validators on first use instead of at import (defer_build),
  through one shared private base class. First-use builds are
  serialised behind a single process-wide lock, since released pydantic
  does not make concurrent first use of a deferred model thread-safe;
  this also fixes a pre-existing concurrent-first-use failure that
  reproduces on main.
- `import mcp` binds the client/server names lazily on first attribute
  access (PEP 562) instead of importing both stacks eagerly, and the
  client no longer imports the server, so client entry points stop
  loading the server, the web stack, httpx2 and cryptography.
- The web application stack (starlette's app machinery, sse_starlette,
  uvicorn) loads with the app builders that use it, and each protocol
  version's wire package loads on the first message parsed for that
  version rather than both loading at import.

On the fresh-interpreter harness `import mcp` is ~0.4x of v1 (main is
~1.6x), the client entry points ~0.6x of v1, `import mcp.server.mcpserver`
~0.7x, and time-to-ready / stdio cold start land at parity with v1. RSS
after `import mcp` is 19 MiB (v1 43.5, main 57). Steady-state per-call
latency is unchanged.

Observable-but-incidental differences (removed incidental namespace
bindings, deeper submodules no longer imported as a side effect of a
bare `import mcp`, get_type_hints needing localns= for a documented set
of callables, pre-first-use introspection) are catalogued in
docs/migration.md; ratchet tests pin the import footprints and the
concurrent-first-use safety.
2026-08-03 15:06:11 +00:00

986 lines
42 KiB
Python

"""Tests for the unified Client class."""
from __future__ import annotations
import contextvars
from collections.abc import AsyncIterator, Iterator
from contextlib import asynccontextmanager, contextmanager
from unittest.mock import patch
import anyio
import mcp_types as types
import pytest
from inline_snapshot import snapshot
from mcp_types import (
CallToolResult,
EmptyResult,
GetPromptResult,
ListPromptsResult,
ListResourcesResult,
ListResourceTemplatesResult,
ListToolsResult,
Prompt,
PromptArgument,
PromptMessage,
PromptsCapability,
ReadResourceResult,
Resource,
ResourcesCapability,
ServerCapabilities,
TextContent,
TextResourceContents,
Tool,
ToolsCapability,
)
from mcp_types.version import LATEST_HANDSHAKE_VERSION
from pydantic import FileUrl
from mcp import MCPDeprecationWarning, MCPError
from mcp.client._memory import InMemoryTransport
from mcp.client._transport import TransportStreams
from mcp.client.client import Client
from mcp.client.session import ClientRequestContext
from mcp.client.streamable_http import streamable_http_client
from mcp.server import Server, ServerRequestContext
from mcp.server.mcpserver import Context, MCPServer
from mcp.shared.memory import MessageStream, create_client_server_memory_streams
from mcp.shared.message import SessionMessage
from tests.interaction._connect import BASE_URL, mounted_app
pytestmark = pytest.mark.anyio
@pytest.fixture
def simple_server() -> Server:
"""Create a simple MCP server for testing."""
async def handle_list_resources(
ctx: ServerRequestContext, params: types.PaginatedRequestParams | None
) -> ListResourcesResult:
return ListResourcesResult(
resources=[Resource(uri="memory://test", name="Test Resource", description="A test resource")]
)
async def handle_subscribe_resource(ctx: ServerRequestContext, params: types.SubscribeRequestParams) -> EmptyResult:
return EmptyResult()
async def handle_unsubscribe_resource(
ctx: ServerRequestContext, params: types.UnsubscribeRequestParams
) -> EmptyResult:
return EmptyResult()
async def handle_set_logging_level(ctx: ServerRequestContext, params: types.SetLevelRequestParams) -> EmptyResult:
return EmptyResult()
async def handle_completion(ctx: ServerRequestContext, params: types.CompleteRequestParams) -> types.CompleteResult:
return types.CompleteResult(completion=types.Completion(values=[]))
return Server( # pyright: ignore[reportDeprecated]
name="test_server",
on_list_resources=handle_list_resources,
on_subscribe_resource=handle_subscribe_resource,
on_unsubscribe_resource=handle_unsubscribe_resource,
on_set_logging_level=handle_set_logging_level,
on_completion=handle_completion,
)
@pytest.fixture
def app() -> MCPServer:
"""Create an MCPServer server for testing."""
server = MCPServer("test")
@server.tool()
def greet(name: str) -> str:
"""Greet someone by name."""
return f"Hello, {name}!"
@server.resource("test://resource")
def test_resource() -> str:
"""A test resource."""
return "Test content"
@server.prompt()
def greeting_prompt(name: str) -> str:
"""A greeting prompt."""
return f"Please greet {name} warmly."
return server
async def test_client_is_initialized(app: MCPServer):
"""Test that the client is initialized after entering context."""
async with Client(app, mode="legacy") as client:
assert client.server_capabilities == snapshot(
ServerCapabilities(
experimental={},
prompts=PromptsCapability(list_changed=False),
resources=ResourcesCapability(subscribe=False, list_changed=False),
tools=ToolsCapability(list_changed=False),
)
)
assert client.server_info is not None
assert client.server_info.name == "test"
async def test_client_exposes_negotiated_protocol_version(app: MCPServer):
"""The negotiated protocol version is readable after initialization."""
async with Client(app, mode="legacy") as client:
assert client.protocol_version == LATEST_HANDSHAKE_VERSION
async def test_client_with_simple_server(simple_server: Server):
"""Test that from_server works with a basic Server instance."""
async with Client(simple_server) as client:
resources = await client.list_resources()
assert resources == snapshot(
ListResourcesResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test_server", "version": ""}},
resources=[Resource(name="Test Resource", uri="memory://test", description="A test resource")],
)
)
async def test_client_send_ping(app: MCPServer):
async with Client(app, mode="legacy") as client:
result = await client.send_ping() # pyright: ignore[reportDeprecated]
assert result == snapshot(EmptyResult())
async def test_client_list_tools(app: MCPServer):
async with Client(app) as client:
result = await client.list_tools()
assert result == snapshot(
ListToolsResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test", "version": ""}},
tools=[
Tool(
name="greet",
description="Greet someone by name.",
input_schema={
"properties": {"name": {"title": "Name", "type": "string"}},
"required": ["name"],
"title": "greetArguments",
"type": "object",
},
output_schema={
"properties": {"result": {"title": "Result", "type": "string"}},
"required": ["result"],
"title": "greetOutput",
"type": "object",
},
)
],
)
)
async def test_client_call_tool(app: MCPServer):
async with Client(app) as client:
result = await client.call_tool("greet", {"name": "World"})
assert result == snapshot(
CallToolResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test", "version": ""}},
content=[TextContent(text="Hello, World!")],
structured_content={"result": "Hello, World!"},
)
)
async def test_read_resource(app: MCPServer):
"""Test reading a resource."""
async with Client(app) as client:
result = await client.read_resource("test://resource")
assert result == snapshot(
ReadResourceResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test", "version": ""}},
contents=[TextResourceContents(uri="test://resource", mime_type="text/plain", text="Test content")],
)
)
async def test_read_resource_error_propagates():
"""MCPError raised by a server handler propagates to the client with its code intact."""
async def handle_read_resource(
ctx: ServerRequestContext, params: types.ReadResourceRequestParams
) -> ReadResourceResult:
raise MCPError(code=404, message="no resource with that URI was found")
server = Server("test", on_read_resource=handle_read_resource)
async with Client(server) as client:
with pytest.raises(MCPError) as exc_info:
await client.read_resource("unknown://example")
assert exc_info.value.error.code == 404
async def test_raise_exceptions_propagates_handler_error_on_modern_inproc_path():
"""`raise_exceptions=True` on the modern in-process path: an unmapped handler
exception reaches the client with its original type chained, instead of being
sanitized to an opaque `INTERNAL_ERROR`."""
async def handle_call_tool(ctx: ServerRequestContext, params: types.CallToolRequestParams) -> CallToolResult:
raise ValueError("boom")
server = Server("test", on_call_tool=handle_call_tool)
async with Client(server, mode="2026-07-28", raise_exceptions=True) as client:
with pytest.raises(MCPError) as exc_info:
await client.call_tool("explode", {})
# The original exception is chained — not swallowed into a generic "Internal server error".
assert isinstance(exc_info.value.__cause__, ValueError)
assert str(exc_info.value.__cause__) == "boom"
async def test_raise_exceptions_false_sanitizes_handler_error_on_modern_inproc_path():
"""`raise_exceptions=False` (the default) on the modern in-process path: an
unmapped handler exception is sanitized to an opaque `INTERNAL_ERROR` so the
in-process path matches the wire path's leak guard."""
async def handle_call_tool(ctx: ServerRequestContext, params: types.CallToolRequestParams) -> CallToolResult:
raise ValueError("boom")
server = Server("test", on_call_tool=handle_call_tool)
async with Client(server, mode="2026-07-28", raise_exceptions=False) as client:
with pytest.raises(MCPError) as exc_info:
await client.call_tool("explode", {})
assert exc_info.value.error.code == types.INTERNAL_ERROR
assert exc_info.value.error.message == "Internal server error"
assert exc_info.value.__cause__ is None
async def test_modern_inproc_path_refuses_server_initiated_requests():
"""The in-process modern entry enforces the same prohibition as the other
modern entries: a handler's request-scoped server-initiated request is
refused server-side with the no-back-channel contract, instead of the
protocol-forbidden frame being delivered to the client."""
async def handle_call_tool(ctx: ServerRequestContext, params: types.CallToolRequestParams) -> CallToolResult:
schema = types.ElicitRequestedSchema(type="object", properties={"x": {"type": "string"}})
await ctx.session.elicit_form("question", schema, related_request_id=ctx.request_id)
raise AssertionError("unreachable: elicit_form must refuse") # pragma: no cover
server = Server("test", on_call_tool=handle_call_tool)
async with Client(server, mode="2026-07-28") as client:
with pytest.raises(MCPError) as exc_info:
await client.call_tool("asker", {})
assert exc_info.value.error.code == types.INVALID_REQUEST
assert "no back-channel" in exc_info.value.error.message
assert "elicitation/create" in exc_info.value.error.message
async def test_get_prompt(app: MCPServer):
"""Test getting a prompt."""
async with Client(app) as client:
result = await client.get_prompt("greeting_prompt", {"name": "Alice"})
assert result == snapshot(
GetPromptResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test", "version": ""}},
description="A greeting prompt.",
messages=[PromptMessage(role="user", content=TextContent(text="Please greet Alice warmly."))],
)
)
def test_client_session_property_before_enter(app: MCPServer):
"""Test that accessing session before context manager raises RuntimeError."""
client = Client(app)
with pytest.raises(RuntimeError, match="Client must be used within an async context manager"):
client.session
async def test_client_reentry_raises_runtime_error(app: MCPServer):
"""Test that reentering a client raises RuntimeError."""
async with Client(app) as client:
with pytest.raises(RuntimeError, match="Client is already entered"):
await client.__aenter__()
async def test_client_send_progress_notification():
"""Test sending progress notification."""
received_from_client = None
event = anyio.Event()
async def handle_progress(ctx: ServerRequestContext, params: types.ProgressNotificationParams) -> None:
nonlocal received_from_client
received_from_client = {"progress_token": params.progress_token, "progress": params.progress}
event.set()
server = Server(name="test_server", on_progress=handle_progress) # pyright: ignore[reportDeprecated]
with anyio.fail_after(5):
async with Client(server, mode="legacy") as client:
await client.send_progress_notification(progress_token="token123", progress=50.0) # pyright: ignore[reportDeprecated]
await event.wait()
assert received_from_client == snapshot({"progress_token": "token123", "progress": 50.0})
async def test_client_subscribe_resource(simple_server: Server):
async with Client(simple_server, mode="legacy") as client:
with pytest.warns(MCPDeprecationWarning, match="use Client.listen"):
result = await client.subscribe_resource("memory://test") # pyright: ignore[reportDeprecated]
assert result == snapshot(EmptyResult())
async def test_client_unsubscribe_resource(simple_server: Server):
async with Client(simple_server, mode="legacy") as client:
with pytest.warns(MCPDeprecationWarning, match="use Client.listen"):
result = await client.unsubscribe_resource("memory://test") # pyright: ignore[reportDeprecated]
assert result == snapshot(EmptyResult())
async def test_client_set_logging_level(simple_server: Server):
"""Test setting logging level."""
async with Client(simple_server, mode="legacy") as client:
result = await client.set_logging_level("debug") # pyright: ignore[reportDeprecated]
assert result == snapshot(EmptyResult())
async def test_client_list_resources_with_params(app: MCPServer):
"""Test listing resources with params parameter."""
async with Client(app) as client:
result = await client.list_resources()
assert result == snapshot(
ListResourcesResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test", "version": ""}},
resources=[
Resource(
name="test_resource",
uri="test://resource",
description="A test resource.",
mime_type="text/plain",
)
],
)
)
async def test_client_list_resource_templates(app: MCPServer):
"""Test listing resource templates with params parameter."""
async with Client(app) as client:
result = await client.list_resource_templates()
assert result == snapshot(
ListResourceTemplatesResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test", "version": ""}}, resource_templates=[]
)
)
async def test_list_prompts(app: MCPServer):
"""Test listing prompts with params parameter."""
async with Client(app) as client:
result = await client.list_prompts()
assert result == snapshot(
ListPromptsResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test", "version": ""}},
prompts=[
Prompt(
name="greeting_prompt",
description="A greeting prompt.",
arguments=[PromptArgument(name="name", required=True)],
)
],
)
)
async def test_complete_with_prompt_reference(simple_server: Server):
"""Test getting completions for a prompt argument."""
async with Client(simple_server) as client:
ref = types.PromptReference(type="ref/prompt", name="test_prompt")
result = await client.complete(ref=ref, argument={"name": "arg", "value": "test"})
assert result == snapshot(
types.CompleteResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test_server", "version": ""}},
completion=types.Completion(values=[]),
)
)
def test_client_with_url_initializes_streamable_http_transport():
with patch("mcp.client.streamable_http.streamable_http_client") as mock:
_ = Client("http://localhost:8000/mcp")
mock.assert_called_once_with("http://localhost:8000/mcp")
async def test_client_uses_transport_directly(app: MCPServer):
transport = InMemoryTransport(app)
async with Client(transport, mode="legacy") as client:
result = await client.call_tool("greet", {"name": "Transport"})
assert result == snapshot(
CallToolResult(
content=[TextContent(text="Hello, Transport!")],
structured_content={"result": "Hello, Transport!"},
)
)
_TEST_CONTEXTVAR = contextvars.ContextVar("test_var", default="initial")
@contextmanager
def _set_test_contextvar(value: str) -> Iterator[None]:
token = _TEST_CONTEXTVAR.set(value)
try:
yield
finally:
_TEST_CONTEXTVAR.reset(token)
async def test_context_propagation():
"""Sender's contextvars.Context is propagated to the server handler."""
server = MCPServer("test")
@server.tool()
async def check_context() -> str:
"""Return the contextvar value visible to the handler."""
return _TEST_CONTEXTVAR.get()
async with Client(server) as client:
with _set_test_contextvar("client_value"):
result = await client.call_tool("check_context", {})
assert result.content[0].text == "client_value", ( # type: ignore[union-attr]
"Server handler did not see the sender's contextvars.Context"
)
async def test_client_auto_mode_probes_discover_then_adopts(simple_server: Server) -> None:
"""`mode='auto'` over an in-process HTTP transport: the `server/discover` probe
reaches the modern entry and the negotiated protocol version is adopted without
an `initialize` handshake."""
with anyio.fail_after(5):
async with (
mounted_app(simple_server) as (http, _),
Client(streamable_http_client(f"{BASE_URL}/mcp", http_client=http), mode="auto") as client,
):
assert client.protocol_version == "2026-07-28"
assert (await client.list_resources()).resources[0].name == "Test Resource"
@asynccontextmanager
async def _stream_loop_transport(server: Server) -> AsyncIterator[TransportStreams]:
"""A Transport whose far end is `Server.run` over crossed memory streams - the stdio shape, in process."""
async with (
create_client_server_memory_streams() as ((client_read, client_write), (server_read, server_write)),
anyio.create_task_group() as tg,
):
tg.start_soon(server.run, server_read, server_write, server.create_initialization_options())
yield client_read, client_write
tg.cancel_scope.cancel()
async def test_client_auto_mode_negotiates_modern_over_a_stream_loop(simple_server: Server) -> None:
"""`mode='auto'` against a real `Server.run` stream loop: the probe reaches the
dual-era driver, the connection locks modern, and feature requests are served
at 2026-07-28 with no `initialize` handshake."""
with anyio.fail_after(5):
async with Client(_stream_loop_transport(simple_server), mode="auto") as client:
assert client.protocol_version == "2026-07-28"
assert (await client.list_resources()).resources[0].name == "Test Resource"
async def test_client_pinned_modern_mode_works_over_a_stream_loop(simple_server: Server) -> None:
"""A pinned-modern client sends no probe: its first envelope-bearing request
locks the stream-loop connection modern and is served."""
with anyio.fail_after(5):
async with Client(_stream_loop_transport(simple_server), mode="2026-07-28") as client:
assert client.protocol_version == "2026-07-28"
assert (await client.list_resources()).resources[0].name == "Test Resource"
async def test_client_legacy_mode_still_handshakes_over_a_stream_loop(simple_server: Server) -> None:
"""`mode='legacy'` against the dual-era stream loop is byte-identical legacy:
the handshake runs and the session lands at a handshake-era version."""
with anyio.fail_after(5):
async with Client(_stream_loop_transport(simple_server), mode="legacy") as client:
assert client.protocol_version == LATEST_HANDSHAKE_VERSION
assert (await client.list_resources()).resources[0].name == "Test Resource"
async def test_client_auto_mode_recovers_from_a_timed_out_probe_over_a_stream_loop(
simple_server: Server, monkeypatch: pytest.MonkeyPatch
) -> None:
"""A probe that outlives the client's discover timeout still succeeds on the
(slow-starting) server and locks the connection modern; the fallback
handshake's -32022 is modern evidence, so one corrective re-probe completes
the connect instead of stranding `mode='auto'`."""
monkeypatch.setattr("mcp.client.session.DISCOVER_TIMEOUT_SECONDS", 0.05)
c2relay_send, c2relay_recv = anyio.create_memory_object_stream[SessionMessage | Exception](32)
relay2s_send, relay2s_recv = anyio.create_memory_object_stream[SessionMessage | Exception](32)
s2c_send, s2c_recv = anyio.create_memory_object_stream[SessionMessage | Exception](32)
async def relay() -> None:
# Hold the client's first frame (the probe) until its second frame (the
# post-timeout initialize) arrives - the deterministic stand-in for a
# server too slow to answer before the client's discover timeout.
held: SessionMessage | Exception | None = None
first = True
async for item in c2relay_recv:
if first:
held, first = item, False
continue
if held is not None:
await relay2s_send.send(held)
held = None
await relay2s_send.send(item)
@asynccontextmanager
async def transport() -> AsyncIterator[TransportStreams]:
async with c2relay_send, c2relay_recv, relay2s_send, relay2s_recv, s2c_send, s2c_recv:
async with anyio.create_task_group() as tg:
tg.start_soon(simple_server.run, relay2s_recv, s2c_send, simple_server.create_initialization_options())
tg.start_soon(relay)
yield s2c_recv, c2relay_send
tg.cancel_scope.cancel()
with anyio.fail_after(10):
async with Client(transport(), mode="auto") as client:
assert client.protocol_version == "2026-07-28"
assert (await client.list_resources()).resources[0].name == "Test Resource"
@pytest.mark.parametrize("code", [types.METHOD_NOT_FOUND, types.REQUEST_TIMEOUT, types.INTERNAL_ERROR])
async def test_client_auto_mode_falls_back_to_initialize_on_legacy_signal(code: int) -> None:
"""`mode='auto'`: any JSON-RPC error from `server/discover` makes
`Client.__aenter__` run the legacy `initialize()` handshake and land at a
handshake-era protocol version. The denylist policy treats every server-sent
rpc-error as "not modern" — including INTERNAL_ERROR, since a legacy server
may crash on the unknown method before reaching its router. A real `Server`
always implements `server/discover`, so the server side is hand-played."""
methods_seen: list[str] = []
async def scripted_server(streams: MessageStream) -> None:
server_read, server_write = streams
async for message in server_read:
assert isinstance(message, SessionMessage)
frame = message.message
assert isinstance(frame, types.JSONRPCRequest | types.JSONRPCNotification)
methods_seen.append(frame.method)
if isinstance(frame, types.JSONRPCNotification):
continue
if frame.method == "server/discover":
error = types.ErrorData(code=code, message="nope")
await server_write.send(SessionMessage(types.JSONRPCError(jsonrpc="2.0", id=frame.id, error=error)))
elif frame.method == "initialize": # pragma: no branch
result = types.InitializeResult(
protocol_version=LATEST_HANDSHAKE_VERSION,
capabilities=ServerCapabilities(),
server_info=types.Implementation(name="legacy-only", version="0.0.1"),
)
await server_write.send(
SessionMessage(
types.JSONRPCResponse(
jsonrpc="2.0",
id=frame.id,
result=result.model_dump(by_alias=True, mode="json", exclude_none=True),
)
)
)
@asynccontextmanager
async def scripted_transport() -> AsyncIterator[TransportStreams]:
async with (
create_client_server_memory_streams() as ((client_read, client_write), server_streams),
anyio.create_task_group() as tg,
):
tg.start_soon(scripted_server, server_streams)
yield client_read, client_write
tg.cancel_scope.cancel()
with anyio.fail_after(5):
async with Client(scripted_transport(), mode="auto") as client:
assert client.protocol_version == LATEST_HANDSHAKE_VERSION
assert client.server_info is not None
assert client.server_info.name == "legacy-only"
assert methods_seen == ["server/discover", "initialize", "notifications/initialized"]
@pytest.mark.anyio
async def test_modern_list_tools_drops_tools_with_invalid_x_mcp_header_but_legacy_does_not() -> None:
"""At 2026-07-28 the spec requires clients to exclude tools whose `x-mcp-header`
annotation is malformed; handshake-era sessions surface them unchanged. Two
tools are advertised — one valid, one with a non-RFC-9110-token header name —
and the modern client sees only the valid one."""
valid = types.Tool(
name="ok",
input_schema={"type": "object", "properties": {"a": {"type": "string", "x-mcp-header": "Region"}}},
)
bad = types.Tool(
name="dropme",
input_schema={"type": "object", "properties": {"a": {"type": "string", "x-mcp-header": "bad name"}}},
)
async def on_list_tools(
ctx: ServerRequestContext, params: types.PaginatedRequestParams | None
) -> types.ListToolsResult:
return types.ListToolsResult(tools=[valid, bad])
server = Server("test", on_list_tools=on_list_tools)
with anyio.fail_after(5):
async with Client(server) as client:
result = await client.list_tools()
assert [t.name for t in result.tools] == ["ok"]
async with Client(server, mode="legacy") as client:
result = await client.list_tools()
assert [t.name for t in result.tools] == ["ok", "dropme"]
_RETIRED_TOOL = Tool(
name="retired",
input_schema={"type": "object", "properties": {"region": {"type": "string", "x-mcp-header": "Region"}}},
output_schema={"type": "object"},
)
_SURVIVOR_TOOL = Tool(name="survivor", input_schema={"type": "object"})
def _scripted_listing_server(listings: list[ListToolsResult]) -> Server:
"""Serves the given listings in order, one per tools/list request."""
async def on_list_tools(ctx: ServerRequestContext, params: types.PaginatedRequestParams | None) -> ListToolsResult:
return listings.pop(0)
return Server("test", on_list_tools=on_list_tools)
async def test_a_complete_listing_prunes_per_tool_state_for_tools_it_no_longer_contains() -> None:
"""SDK-defined: a complete (uncursored, cursorless) listing is the full tool universe, so the
header map and output schema derived from an earlier listing of a now-absent tool are dropped."""
server = _scripted_listing_server(
[
ListToolsResult(tools=[_RETIRED_TOOL, _SURVIVOR_TOOL]),
ListToolsResult(tools=[_SURVIVOR_TOOL]),
]
)
with anyio.fail_after(5):
async with Client(server) as client:
await client.session.list_tools()
# Compile the retired tool's output-schema validator so its eviction is observable.
await client.session.validate_tool_result("retired", CallToolResult(content=[], structured_content={}))
assert set(client.session._x_mcp_header_maps) == {"retired", "survivor"}
assert set(client.session._tool_output_schemas) == {"retired", "survivor"}
assert set(client.session._tool_output_validators) == {"retired"}
await client.session.list_tools()
assert set(client.session._x_mcp_header_maps) == {"survivor"}
assert set(client.session._tool_output_schemas) == {"survivor"}
assert client.session._tool_output_validators == {}
async def test_a_complete_listing_prunes_output_schemas_on_a_legacy_session_too() -> None:
"""SDK-defined: the prune is era-independent -- legacy sessions cache output schemas the same
way (their header-map dict just stays empty, since the x-mcp-header filter is 2026-only)."""
server = _scripted_listing_server(
[
ListToolsResult(tools=[_RETIRED_TOOL, _SURVIVOR_TOOL]),
ListToolsResult(tools=[_SURVIVOR_TOOL]),
]
)
with anyio.fail_after(5):
async with Client(server, mode="legacy") as client:
await client.session.list_tools()
assert set(client.session._tool_output_schemas) == {"retired", "survivor"}
assert client.session._x_mcp_header_maps == {}
await client.session.list_tools()
assert set(client.session._tool_output_schemas) == {"survivor"}
async def test_a_listing_with_a_next_cursor_prunes_no_per_tool_state() -> None:
"""SDK-defined: a first page carrying next_cursor is not the full universe -- state for tools
expected on later pages must survive it."""
server = _scripted_listing_server(
[
ListToolsResult(tools=[_RETIRED_TOOL, _SURVIVOR_TOOL]),
ListToolsResult(tools=[_SURVIVOR_TOOL], next_cursor="2"),
]
)
with anyio.fail_after(5):
async with Client(server) as client:
await client.session.list_tools()
await client.session.list_tools()
assert set(client.session._x_mcp_header_maps) == {"retired", "survivor"}
assert set(client.session._tool_output_schemas) == {"retired", "survivor"}
async def test_a_cursor_page_fetch_prunes_no_per_tool_state() -> None:
"""SDK-defined: a continuation page is partial even when it ends the pagination (no
next_cursor) -- only an uncursored single-page listing prunes."""
server = _scripted_listing_server(
[
ListToolsResult(tools=[_RETIRED_TOOL, _SURVIVOR_TOOL]),
ListToolsResult(tools=[_SURVIVOR_TOOL]),
]
)
with anyio.fail_after(5):
async with Client(server) as client:
await client.session.list_tools()
await client.session.list_tools(params=types.PaginatedRequestParams(cursor="2"))
assert set(client.session._x_mcp_header_maps) == {"retired", "survivor"}
assert set(client.session._tool_output_schemas) == {"retired", "survivor"}
def test_client_rejects_handshake_era_mode_at_construction() -> None:
"""A handshake-era protocol-version string passed as `mode=` is rejected by
`__post_init__` with a hint to use `mode='legacy'` — the version-pin path is
modern-only."""
server = MCPServer("test")
with pytest.raises(ValueError, match=r"handshake-era version; use mode='legacy'"):
Client(server, mode="2025-06-18")
with pytest.raises(ValueError, match=r"mode must be 'legacy', 'auto', or one of"):
Client(server, mode="not-a-version")
# ── SEP-2322 multi-round-trip auto-loop ────────────────────────────────────────
_NAME_SCHEMA = {"type": "object", "properties": {"name": {"type": "string"}}, "required": ["name"]}
def _name_elicitation(message: str = "What is your name?") -> types.ElicitRequest:
return types.ElicitRequest(params=types.ElicitRequestFormParams(message=message, requested_schema=_NAME_SCHEMA))
async def test_call_tool_auto_loop_dispatches_elicitation_then_returns_final_result() -> None:
"""When the server returns `InputRequiredResult` carrying an elicitation,
`Client.call_tool` routes it to `elicitation_callback` and retries
automatically — the caller sees only the terminal `CallToolResult`."""
server = MCPServer("test")
@server.tool()
async def greet(ctx: Context) -> str | types.InputRequiredResult:
responses = ctx.input_responses
if responses and "user_name" in responses:
answer = responses["user_name"]
assert isinstance(answer, types.ElicitResult)
assert answer.content is not None
return f"Hello, {answer.content['name']}!"
return types.InputRequiredResult(input_requests={"user_name": _name_elicitation()})
callback_params: list[types.ElicitRequestParams] = []
async def elicitation_callback(
context: ClientRequestContext, params: types.ElicitRequestParams
) -> types.ElicitResult | types.ErrorData:
callback_params.append(params)
assert context.request_id == "user_name" # the inputRequests key is the request id
return types.ElicitResult(action="accept", content={"name": "Ada"})
with anyio.fail_after(5):
async with Client(server, elicitation_callback=elicitation_callback) as client:
result = await client.call_tool("greet")
assert result == snapshot(
CallToolResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test", "version": ""}},
content=[TextContent(text="Hello, Ada!")],
structured_content={"result": "Hello, Ada!"},
)
)
assert len(callback_params) == 1
assert isinstance(callback_params[0], types.ElicitRequestFormParams)
assert callback_params[0].message == "What is your name?"
assert callback_params[0].requested_schema == _NAME_SCHEMA
async def test_call_tool_auto_loop_dispatches_sampling_then_returns_final_result() -> None:
"""`InputRequiredResult` with an embedded `CreateMessageRequest` is routed
to `sampling_callback` and the call retried with the model's reply."""
server = MCPServer("test")
@server.tool()
async def ask(ctx: Context) -> str | types.InputRequiredResult:
responses = ctx.input_responses
if responses and "q" in responses:
answer = responses["q"]
assert isinstance(answer, types.CreateMessageResult)
assert answer.content.type == "text"
return f"Model said: {answer.content.text}"
return types.InputRequiredResult(
input_requests={
"q": types.CreateMessageRequest(
params=types.CreateMessageRequestParams(
messages=[types.SamplingMessage(role="user", content=TextContent(text="Capital of France?"))],
max_tokens=10,
)
)
}
)
callback_params: list[types.CreateMessageRequestParams] = []
async def sampling_callback(
context: ClientRequestContext, params: types.CreateMessageRequestParams
) -> types.CreateMessageResult | types.ErrorData:
callback_params.append(params)
return types.CreateMessageResult(role="assistant", content=TextContent(text="Paris"), model="echo")
with anyio.fail_after(5):
async with Client(server, sampling_callback=sampling_callback) as client:
result = await client.call_tool("ask")
assert result == snapshot(
CallToolResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test", "version": ""}},
content=[TextContent(text="Model said: Paris")],
structured_content={"result": "Model said: Paris"},
)
)
assert len(callback_params) == 1
assert callback_params[0].messages[0].content == TextContent(text="Capital of France?")
async def test_call_tool_auto_loop_dispatches_list_roots_then_returns_final_result() -> None:
"""`InputRequiredResult` with an embedded `ListRootsRequest` is routed to
`list_roots_callback` and the call retried with the returned roots."""
server = MCPServer("test")
@server.tool()
async def count_roots(ctx: Context) -> str | types.InputRequiredResult:
responses = ctx.input_responses
if responses and "roots" in responses:
answer = responses["roots"]
assert isinstance(answer, types.ListRootsResult)
return f"Client exposed {len(answer.roots)} root(s)."
return types.InputRequiredResult(input_requests={"roots": types.ListRootsRequest()})
callback_called: list[ClientRequestContext] = []
async def list_roots_callback(context: ClientRequestContext) -> types.ListRootsResult | types.ErrorData:
callback_called.append(context)
return types.ListRootsResult(roots=[types.Root(uri=FileUrl("file:///workspace"))])
with anyio.fail_after(5):
async with Client(server, list_roots_callback=list_roots_callback) as client:
result = await client.call_tool("count_roots")
assert result == snapshot(
CallToolResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test", "version": ""}},
content=[TextContent(text="Client exposed 1 root(s).")],
structured_content={"result": "Client exposed 1 root(s)."},
)
)
assert len(callback_called) == 1
assert callback_called[0].request_id == "roots"
async def test_call_tool_auto_loop_round_trips_evolving_request_state_across_three_rounds() -> None:
"""A three-round flow where each `InputRequiredResult.request_state`
encodes the round number: the driver echoes it back byte-exact, the server
advances per round, and the elicitation callback runs once per round."""
server = MCPServer("test")
@server.tool()
async def multi(ctx: Context) -> str | types.InputRequiredResult:
# Round number is the integer the server stashed in `request_state` last leg.
round_num = int(ctx.request_state) if ctx.request_state else 0
if round_num == 3:
return "done after 3 rounds"
next_round = round_num + 1
return types.InputRequiredResult(
input_requests={f"step{next_round}": _name_elicitation(f"Round {next_round}?")},
request_state=str(next_round),
)
messages: list[str] = []
async def elicitation_callback(
context: ClientRequestContext, params: types.ElicitRequestParams
) -> types.ElicitResult | types.ErrorData:
assert isinstance(params, types.ElicitRequestFormParams)
messages.append(params.message)
return types.ElicitResult(action="accept", content={"name": "x"})
with anyio.fail_after(5):
async with Client(server, elicitation_callback=elicitation_callback) as client:
result = await client.call_tool("multi")
assert result.content == [TextContent(text="done after 3 rounds")]
assert messages == ["Round 1?", "Round 2?", "Round 3?"]
async def test_call_tool_auto_loop_raises_mcp_error_when_no_callback_registered() -> None:
"""SDK-defined: with no `elicitation_callback`, the default returns
`ErrorData(INVALID_REQUEST, ...)` and the driver raises it as `MCPError`
rather than retrying."""
server = MCPServer("test")
@server.tool()
async def needs_input(ctx: Context) -> str | types.InputRequiredResult:
if ctx.input_responses:
raise NotImplementedError # unreachable: client errors before retrying
return types.InputRequiredResult(input_requests={"ask": _name_elicitation()})
async with Client(server) as client:
with anyio.fail_after(5), pytest.raises(MCPError) as exc:
await client.call_tool("needs_input")
assert exc.value.error.code == types.INVALID_REQUEST
async def test_get_prompt_auto_loop_resolves_input_required_via_callbacks() -> None:
"""`Client.get_prompt` runs the same driver as `call_tool`: an
`InputRequiredResult` from `prompts/get` is fulfilled and retried."""
async def handler(
ctx: ServerRequestContext, params: types.GetPromptRequestParams
) -> types.GetPromptResult | types.InputRequiredResult:
assert params.name == "summary"
if params.input_responses and "ask" in params.input_responses:
return GetPromptResult(messages=[PromptMessage(role="user", content=TextContent(text="ok"))])
return types.InputRequiredResult(input_requests={"ask": _name_elicitation()})
server = Server("test")
server.add_request_handler("prompts/get", types.GetPromptRequestParams, handler)
async def elicitation_callback(
context: ClientRequestContext, params: types.ElicitRequestParams
) -> types.ElicitResult | types.ErrorData:
return types.ElicitResult(action="accept", content={"name": "x"})
with anyio.fail_after(5):
async with Client(server, mode="2026-07-28", elicitation_callback=elicitation_callback) as client:
result = await client.get_prompt("summary")
assert result == snapshot(
GetPromptResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test", "version": ""}},
messages=[PromptMessage(role="user", content=TextContent(text="ok"))],
)
)
async def test_read_resource_auto_loop_resolves_input_required_via_callbacks() -> None:
"""`Client.read_resource` runs the same driver as `call_tool`: an
`InputRequiredResult` from `resources/read` is fulfilled and retried."""
async def handler(
ctx: ServerRequestContext, params: types.ReadResourceRequestParams
) -> types.ReadResourceResult | types.InputRequiredResult:
assert params.uri == "memory://gated"
if params.input_responses and "ask" in params.input_responses:
return ReadResourceResult(contents=[TextResourceContents(uri="memory://gated", text="unlocked")])
return types.InputRequiredResult(input_requests={"ask": _name_elicitation()})
server = Server("test")
server.add_request_handler("resources/read", types.ReadResourceRequestParams, handler)
async def elicitation_callback(
context: ClientRequestContext, params: types.ElicitRequestParams
) -> types.ElicitResult | types.ErrorData:
return types.ElicitResult(action="accept", content={"name": "x"})
with anyio.fail_after(5):
async with Client(server, mode="2026-07-28", elicitation_callback=elicitation_callback) as client:
result = await client.read_resource("memory://gated")
assert result == snapshot(
ReadResourceResult(
_meta={"io.modelcontextprotocol/serverInfo": {"name": "test", "version": ""}},
contents=[TextResourceContents(uri="memory://gated", text="unlocked")],
)
)