forked from modelcontextprotocol/python-sdk
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathtest_lowlevel_input_validation.py
More file actions
416 lines (340 loc) · 14.6 KB
/
test_lowlevel_input_validation.py
File metadata and controls
416 lines (340 loc) · 14.6 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
"""Test input schema validation for lowlevel server."""
import logging
from collections.abc import Awaitable, Callable
from typing import Any
import anyio
import pytest
from mcp.client.session import ClientSession
from mcp.server import Server
from mcp.server.lowlevel import NotificationOptions
from mcp.server.models import InitializationOptions
from mcp.server.session import ServerSession
from mcp.shared.exceptions import McpError
from mcp.shared.message import SessionMessage
from mcp.shared.session import RequestResponder
from mcp.types import (
METHOD_NOT_FOUND,
CallToolResult,
ClientResult,
ErrorData,
ServerNotification,
ServerRequest,
TextContent,
Tool,
)
async def run_tool_test(
tools: list[Tool],
call_tool_handler: Callable[[str, dict[str, Any]], Awaitable[list[TextContent]]],
test_callback: Callable[[ClientSession], Awaitable[Any]],
) -> Any:
"""Helper to run a tool test with minimal boilerplate.
Args:
tools: List of tools to register
call_tool_handler: Handler function for tool calls
test_callback: Async function that performs the test using the client session
Returns:
The result of the tool call
"""
server = Server("test")
@server.list_tools()
async def list_tools():
return tools
@server.call_tool()
async def call_tool(name: str, arguments: dict[str, Any]) -> list[TextContent]:
return await call_tool_handler(name, arguments)
server_to_client_send, server_to_client_receive = anyio.create_memory_object_stream[SessionMessage](10)
client_to_server_send, client_to_server_receive = anyio.create_memory_object_stream[SessionMessage](10)
# Message handler for client
async def message_handler(
message: RequestResponder[ServerRequest, ClientResult] | ServerNotification | Exception,
) -> None:
if isinstance(message, Exception):
raise message
# Server task
async def run_server():
async with ServerSession(
client_to_server_receive,
server_to_client_send,
InitializationOptions(
server_name="test-server",
server_version="1.0.0",
capabilities=server.get_capabilities(
notification_options=NotificationOptions(),
experimental_capabilities={},
),
),
) as server_session:
async with anyio.create_task_group() as tg:
async def handle_messages():
async for message in server_session.incoming_messages:
await server._handle_message(message, server_session, {}, False)
tg.start_soon(handle_messages)
await anyio.sleep_forever()
# Run the test
async with anyio.create_task_group() as tg:
tg.start_soon(run_server)
async with ClientSession(
server_to_client_receive,
client_to_server_send,
message_handler=message_handler,
) as client_session:
# Initialize the session
await client_session.initialize()
# Run the test callback
result = await test_callback(client_session)
# Cancel the server task
tg.cancel_scope.cancel()
return result
def create_add_tool() -> Tool:
"""Create a standard 'add' tool for testing."""
return Tool(
name="add",
description="Add two numbers",
inputSchema={
"type": "object",
"properties": {
"a": {"type": "number"},
"b": {"type": "number"},
},
"required": ["a", "b"],
"additionalProperties": False,
},
)
@pytest.mark.anyio
async def test_valid_tool_call():
"""Test that valid arguments pass validation."""
async def call_tool_handler(name: str, arguments: dict[str, Any]) -> list[TextContent]:
if name == "add":
result = arguments["a"] + arguments["b"]
return [TextContent(type="text", text=f"Result: {result}")]
else:
raise ValueError(f"Unknown tool: {name}")
async def test_callback(client_session: ClientSession) -> CallToolResult:
return await client_session.call_tool("add", {"a": 5, "b": 3})
result = await run_tool_test([create_add_tool()], call_tool_handler, test_callback)
# Verify results
assert result is not None
assert not result.isError
assert len(result.content) == 1
assert result.content[0].type == "text"
assert isinstance(result.content[0], TextContent)
assert result.content[0].text == "Result: 8"
@pytest.mark.anyio
async def test_invalid_tool_call_missing_required():
"""Test that missing required arguments fail validation."""
async def call_tool_handler(name: str, arguments: dict[str, Any]) -> list[TextContent]:
# This should not be reached due to validation
raise RuntimeError("Should not reach here")
async def test_callback(client_session: ClientSession) -> CallToolResult:
return await client_session.call_tool("add", {"a": 5}) # missing 'b'
result = await run_tool_test([create_add_tool()], call_tool_handler, test_callback)
# Verify results
assert result is not None
assert result.isError
assert len(result.content) == 1
assert result.content[0].type == "text"
assert isinstance(result.content[0], TextContent)
assert "Input validation error" in result.content[0].text
assert "'b' is a required property" in result.content[0].text
@pytest.mark.anyio
async def test_invalid_tool_call_wrong_type():
"""Test that wrong argument types fail validation."""
async def call_tool_handler(name: str, arguments: dict[str, Any]) -> list[TextContent]:
# This should not be reached due to validation
raise RuntimeError("Should not reach here")
async def test_callback(client_session: ClientSession) -> CallToolResult:
return await client_session.call_tool("add", {"a": "five", "b": 3}) # 'a' should be number
result = await run_tool_test([create_add_tool()], call_tool_handler, test_callback)
# Verify results
assert result is not None
assert result.isError
assert len(result.content) == 1
assert result.content[0].type == "text"
assert isinstance(result.content[0], TextContent)
assert "Input validation error" in result.content[0].text
assert "'five' is not of type 'number'" in result.content[0].text
@pytest.mark.anyio
async def test_cache_refresh_on_missing_tool():
"""Test that tool cache is refreshed when tool is not found."""
tools = [
Tool(
name="multiply",
description="Multiply two numbers",
inputSchema={
"type": "object",
"properties": {
"x": {"type": "number"},
"y": {"type": "number"},
},
"required": ["x", "y"],
},
)
]
async def call_tool_handler(name: str, arguments: dict[str, Any]) -> list[TextContent]:
if name == "multiply":
result = arguments["x"] * arguments["y"]
return [TextContent(type="text", text=f"Result: {result}")]
else:
raise ValueError(f"Unknown tool: {name}")
async def test_callback(client_session: ClientSession) -> CallToolResult:
# Call tool without first listing tools (cache should be empty)
# The cache should be refreshed automatically
return await client_session.call_tool("multiply", {"x": 10, "y": 20})
result = await run_tool_test(tools, call_tool_handler, test_callback)
# Verify results - should work because cache will be refreshed
assert result is not None
assert not result.isError
assert len(result.content) == 1
assert result.content[0].type == "text"
assert isinstance(result.content[0], TextContent)
assert result.content[0].text == "Result: 200"
@pytest.mark.anyio
async def test_enum_constraint_validation():
"""Test that enum constraints are validated."""
tools = [
Tool(
name="greet",
description="Greet someone",
inputSchema={
"type": "object",
"properties": {
"name": {"type": "string"},
"title": {"type": "string", "enum": ["Mr", "Ms", "Dr"]},
},
"required": ["name"],
},
)
]
async def call_tool_handler(name: str, arguments: dict[str, Any]) -> list[TextContent]:
# This should not be reached due to validation failure
raise RuntimeError("Should not reach here")
async def test_callback(client_session: ClientSession) -> CallToolResult:
return await client_session.call_tool("greet", {"name": "Smith", "title": "Prof"}) # Invalid title
result = await run_tool_test(tools, call_tool_handler, test_callback)
# Verify results
assert result is not None
assert result.isError
assert len(result.content) == 1
assert result.content[0].type == "text"
assert isinstance(result.content[0], TextContent)
assert "Input validation error" in result.content[0].text
assert "'Prof' is not one of" in result.content[0].text
@pytest.mark.anyio
async def test_tool_not_in_list_logs_warning_before_protocol_error(caplog):
"""Test that calling a tool not in list_tools logs a warning before returning protocol error."""
tools = [
Tool(
name="add",
description="Add two numbers",
inputSchema={
"type": "object",
"properties": {
"a": {"type": "number"},
"b": {"type": "number"},
},
"required": ["a", "b"],
},
)
]
async def call_tool_handler(name: str, arguments: dict[str, Any]) -> list[TextContent]:
# This should not be reached due to protocol error for unknown tools
if name == "add":
result = arguments["a"] + arguments["b"]
return [TextContent(type="text", text=f"Result: {result}")]
else:
raise ValueError(f"Unknown tool: {name}")
async def test_callback(client_session: ClientSession):
# Call a tool that's not in the list - should now raise McpError
try:
return await client_session.call_tool("unknown_tool", {"invalid": "args"})
except McpError as e:
return e
with caplog.at_level(logging.WARNING):
result = await run_tool_test(tools, call_tool_handler, test_callback)
# Verify it's the correct protocol error
assert isinstance(result, McpError), f"Expected McpError but got {type(result)}"
assert isinstance(result.error, ErrorData)
assert result.error.code == METHOD_NOT_FOUND
assert "Unknown tool: unknown_tool" in result.error.message
# Verify warning was still logged during the tool lookup process
assert any(
"Tool 'unknown_tool' not listed, no validation will be performed" in record.message for record in caplog.records
)
@pytest.mark.anyio
async def test_unknown_tool_returns_protocol_error():
"""Test that calling an unknown tool returns a proper JSON-RPC protocol error."""
tools = [
Tool(
name="add",
description="Add two numbers",
inputSchema={
"type": "object",
"properties": {
"a": {"type": "number"},
"b": {"type": "number"},
},
"required": ["a", "b"],
},
)
]
async def call_tool_handler(name: str, arguments: dict[str, Any]) -> list[TextContent]:
# This should not be reached for unknown tools due to protocol error
if name == "add":
result = arguments["a"] + arguments["b"]
return [TextContent(type="text", text=f"Result: {result}")]
else:
raise ValueError(f"Unknown tool: {name}")
async def test_callback(client_session: ClientSession):
# Try to call a tool that doesn't exist - should raise McpError
try:
return await client_session.call_tool("unknown_tool", {"invalid": "args"})
except McpError as e:
return e
result = await run_tool_test(tools, call_tool_handler, test_callback)
# Verify it's the correct protocol error
assert isinstance(result, McpError), f"Expected McpError but got {type(result)}"
assert isinstance(result.error, ErrorData)
assert result.error.code == METHOD_NOT_FOUND
assert "Unknown tool: unknown_tool" in result.error.message
@pytest.mark.anyio
async def test_tool_execution_error_vs_protocol_error():
"""Test the difference between tool execution errors and protocol errors."""
tools = [
Tool(
name="failing_tool",
description="A tool that always fails during execution",
inputSchema={
"type": "object",
"properties": {
"input": {"type": "string"},
},
},
)
]
async def call_tool_handler(name: str, arguments: dict[str, Any]) -> list[TextContent]:
if name == "failing_tool":
# This should cause a tool execution error (not a protocol error)
raise RuntimeError("Tool execution failed")
else:
raise ValueError(f"Unknown tool: {name}")
# Test 1: Tool execution error (valid tool that fails)
async def test_execution_error(client_session: ClientSession):
return await client_session.call_tool("failing_tool", {"input": "test"})
execution_result = await run_tool_test(tools, call_tool_handler, test_execution_error)
# Should return CallToolResult with isError=True (tool execution error)
assert isinstance(execution_result, CallToolResult)
assert execution_result.isError
assert isinstance(execution_result.content[0], TextContent)
assert "Tool execution failed" in execution_result.content[0].text
# Test 2: Protocol error (unknown tool)
async def test_protocol_error(client_session: ClientSession):
try:
return await client_session.call_tool("nonexistent_tool", {"input": "test"})
except McpError as e:
return e
protocol_result = await run_tool_test(tools, call_tool_handler, test_protocol_error)
# Should return McpError (protocol error)
assert isinstance(protocol_result, McpError), f"Expected McpError but got {type(protocol_result)}"
assert isinstance(protocol_result.error, ErrorData)
assert protocol_result.error.code == METHOD_NOT_FOUND
assert "Unknown tool: nonexistent_tool" in protocol_result.error.message