Support sampling/createMessage per MCP specification

## Motivation and Context

The MCP specification defines `sampling/createMessage` to request LLM completions from the client.
The Ruby SDK did not yet support server-to-client requests,
so the server had no way to ask the client to sample an LLM.
This caused the `tools-call-sampling` conformance test to fail.

This adds `Server#create_sampling_message` to send a `sampling/createMessage` request
to the client via the transport layer. The method validates that the client declared
the `sampling` capability during initialization, and optionally checks the
`sampling.tools` capability when tools or tool_choice parameters are provided.

On the transport side, `Transport#send_request` is introduced as an abstract method,
with implementations in both `StdioTransport` and `StreamableHTTPTransport`.
Each transport assigns a unique request ID, sends the JSON-RPC request to the client,
and blocks until the client responds.

Client capabilities are stored per session via `ServerSession`, so validation is scoped
to the originating client. `ServerSession#create_sampling_message` uses
`build_sampling_params` for capability validation and sends the request directly through
the transport with the session ID, following the same pattern as `notify_progress`.

`Server#create_sampling_message` is the public API for single-client transports
(e.g., StdioTransport).

For multi-client transports (e.g., StreamableHTTPTransport),
`ServerSession#create_sampling_message` must be used to route requests to the correct client.
Tools using `server_context.create_sampling_message` automatically route to the correct session
regardless of transport.

`StreamableHTTPTransport#send_request` requires `session_id` to prevent broadcasting sampling requests
to all connected clients. Missing or invalid session IDs raise explicit errors instead of silently failing.

The `include_context` parameter is soft-deprecated in the MCP specification but is included
for compatibility with existing clients that declare the `sampling.context` capability.
Python and TypeScript SDKs also retain this parameter.

Ref: https://modelcontextprotocol.io/specification/latest/server/utilities/sampling

## How Has This Been Tested?

Added comprehensive tests in `test/mcp/server_sampling_test.rb` covering required and optional parameters,
capability validation, error handling, nil-param omission, per-session capability isolation,
and HTTP init capability scoping.

Added transport tests in `test/mcp/server/transports/stdio_transport_test.rb` and
`test/mcp/server/transports/streamable_http_transport_test.rb` for `send_request` round-trip behavior,
error responses, stateless mode rejection, missing session_id, invalid session_id, and no-active-stream error.

Conformance: `tools-call-sampling` is removed from expected failures. The conformance server's `TestSampling`
tool now calls `server_context.create_sampling_message` instead of returning a stub error.

This aligns the Ruby SDK with the MCP specification and other SDK implementations.

## Breaking Changes

None for end users.
`Transport#send_request` is a new abstract method. `ServerSession#create_sampling_message` is a new method.
This commit is contained in:
Koichi ITO 2026-03-29 16:05:10 +09:00
commit da5f9225ac

View file

@ -151,6 +151,247 @@ module MCP
$stdout = original_stdout
end
end
test "send_request sends request to stdout and waits for response" do
reader, writer = IO.pipe
output = StringIO.new
original_stdin = $stdin
original_stdout = $stdout
begin
$stdin = reader
$stdout = output
@transport.instance_variable_set(:@open, true)
# Send response from client in a thread.
Thread.new do
sleep(0.05) # Wait for request to be written to `StringIO`.
request = JSON.parse(output.string.lines.first, symbolize_names: true)
response = {
jsonrpc: "2.0",
id: request[:id],
result: { content: "test response" },
}
writer.puts(response.to_json)
writer.flush
end
result = @transport.send_request("test/method", { param: "value" })
assert_equal({ content: "test response" }, result)
# Verify request was sent.
request = JSON.parse(output.string.lines.first, symbolize_names: true)
assert_equal("2.0", request[:jsonrpc])
assert_equal("test/method", request[:method])
assert_equal({ param: "value" }, request[:params])
assert(request[:id])
ensure
$stdin = original_stdin
$stdout = original_stdout
begin
writer.close
rescue
nil
end
begin
reader.close
rescue
nil
end
end
end
test "send_request raises on error response from client" do
reader, writer = IO.pipe
output = StringIO.new
original_stdin = $stdin
original_stdout = $stdout
begin
$stdin = reader
$stdout = output
@transport.instance_variable_set(:@open, true)
Thread.new do
sleep(0.05) # Wait for request to be written to `StringIO`.
request = JSON.parse(output.string.lines.first, symbolize_names: true)
error_response = {
jsonrpc: "2.0",
id: request[:id],
error: { code: -1, message: "User rejected sampling request" },
}
writer.puts(error_response.to_json)
writer.flush
end
error = assert_raises(StandardError) do
@transport.send_request("sampling/createMessage", { messages: [] })
end
assert_equal("Client returned an error for sampling/createMessage request (code: -1): User rejected sampling request", error.message)
ensure
$stdin = original_stdin
$stdout = original_stdout
begin
writer.close
rescue
nil
end
begin
reader.close
rescue
nil
end
end
end
test "send_request does not double-report intentional raises via exception_reporter" do
reader, writer = IO.pipe
output = StringIO.new
original_stdin = $stdin
original_stdout = $stdout
reported_errors = []
original_reporter = MCP.configuration.exception_reporter
begin
MCP.configuration.exception_reporter = ->(e, ctx) { reported_errors << [e, ctx] }
$stdin = reader
$stdout = output
@transport.instance_variable_set(:@open, true)
Thread.new do
sleep(0.05) # Wait for request to be written to `StringIO`.
request = JSON.parse(output.string.lines.first, symbolize_names: true)
error_response = {
jsonrpc: "2.0",
id: request[:id],
error: { code: -1, message: "rejected" },
}
writer.puts(error_response.to_json)
writer.flush
end
assert_raises(StandardError) do
@transport.send_request("sampling/createMessage", { messages: [] })
end
assert_empty(reported_errors)
ensure
MCP.configuration.exception_reporter = original_reporter
$stdin = original_stdin
$stdout = original_stdout
begin
writer.close
rescue
nil
end
begin
reader.close
rescue
nil
end
end
end
test "send_request processes interleaved requests via session" do
reader, writer = IO.pipe
output = StringIO.new
original_stdin = $stdin
original_stdout = $stdout
begin
$stdin = reader
$stdout = output
@transport.instance_variable_set(:@open, true)
# Initialize a session so @session is set.
session = MCP::ServerSession.new(server: @server, transport: @transport)
@transport.instance_variable_set(:@session, session)
Thread.new do
sleep(0.05) # Wait for request to be written to `StringIO`.
request = JSON.parse(output.string.lines.first, symbolize_names: true)
# Send an interleaved ping request before the response.
ping = { jsonrpc: "2.0", method: "ping", id: "ping-1" }
writer.puts(ping.to_json)
writer.flush
sleep(0.05) # Wait for the ping to be processed.
# Then send the actual response.
response = {
jsonrpc: "2.0",
id: request[:id],
result: { content: "done" },
}
writer.puts(response.to_json)
writer.flush
end
result = @transport.send_request("test/method", { param: "value" })
assert_equal({ content: "done" }, result)
# Verify the interleaved ping was handled (response sent to output).
lines = output.string.lines
ping_response = lines.find { |l| l.include?("ping-1") }
assert(ping_response, "Interleaved ping request should have been handled")
ensure
$stdin = original_stdin
$stdout = original_stdout
begin
writer.close
rescue
nil
end
begin
reader.close
rescue
nil
end
end
end
test "send_request raises when transport is closed while waiting" do
reader, writer = IO.pipe
output = StringIO.new
original_stdin = $stdin
original_stdout = $stdout
begin
$stdin = reader
$stdout = output
@transport.instance_variable_set(:@open, true)
# Close transport while waiting for response.
Thread.new do
sleep(0.05) # Wait for request to be written to `StringIO`.
@transport.instance_variable_set(:@open, false)
writer.close
end
error = assert_raises(RuntimeError) do
@transport.send_request("sampling/createMessage", { messages: [] })
end
assert_equal("Transport closed while waiting for response to sampling/createMessage request.", error.message)
ensure
$stdin = original_stdin
$stdout = original_stdout
begin
writer.close
rescue IOError
nil
end
begin
reader.close
rescue IOError
nil
end
end
end
end
end
end