This suite enumerates the MCP interaction model as end-to-end tests: one test per piece of functionality, asserting the full client↔server round trip through the public API. It exists to pin the SDK's observable behaviour — every request type, every notification direction, every error plane — so that internal rewrites of the send/receive path can be proven equivalent by running the suite before and after.

uv run --frozen pytest tests/interaction/

The whole suite is in-process and event-driven — including the streamable HTTP, SSE, and OAuth flows — with a single subprocess test for stdio.

• Public API only. Tests drive a Client connected to a Server or MCPServer. Nothing reaches into session internals, so the suite keeps working when those internals change. ClientSession is used directly only for behaviours Client cannot express (skipping initialization, requesting a non-default protocol version).
• Pin current behaviour. Every test passes against the current main, including behaviours that diverge from the specification. A failing or xfailed test proves nothing about whether a rewrite preserved behaviour; a passing test that pins the wrong output exactly does. Known divergences are recorded as data on the requirement (see below), not worked around in the test.
• Spec-mandated assertions, not implementation quirks. Error codes are asserted against the constants in mcp_types; error message strings are pinned only where they are the SDK's own deliberate output.
• No sleeps, no real I/O. Concurrency is coordinated with anyio.Event; every wait that could hang is bounded by anyio.fail_after(5). A test that must let in-flight deliveries settle before teardown (an abandoned request's late error response, say) may use anyio.wait_all_tasks_blocked(): the whole suite is single-loop and task-driven, so quiescence is deterministic. The HTTP and OAuth tests drive the Starlette app in-process through the suite's streaming ASGI bridge (transports/_bridge.py), which delivers each response chunk as the server produces it — full duplex, but still no sockets, threads, or subprocesses anywhere outside the one stdio test.

tests/interaction/
  _requirements.py      the requirements manifest (see below)
  _helpers.py           the wire-recording transport
  _connect.py           the transport-parametrized connection factories
  conftest.py           the connect fixture (the transport matrix)
  test_coverage.py      enforces the manifest ↔ test contract
  lowlevel/             one file per feature area, against the low-level Server
  mcpserver/            the same feature areas in MCPServer's natural idiom
  transports/           behaviour specific to one transport (sessions, resumability, framing)
  auth/                 OAuth flows against an in-process authorization server

The two server APIs produce genuinely different wire output for the same conceptual feature (MCPServer generates schemas, converts exceptions to isError results, attaches structured content), so they get parallel directories with mirrored file names rather than one parametrized test body — each directory pins its flavour's true output exactly.

Transport-agnostic tests take the connect fixture instead of constructing Client(server) directly, and therefore run once per transport: over the in-memory transport, over the server's real streamable HTTP app driven in-process through the streaming bridge (in both stateful and stateless configurations), and over the legacy SSE transport the same way. A test connects with async with connect(server, ...) as client: and asserts the same output on every leg, because the transport is not supposed to change observable behaviour. Requirements that need a server-to-client back-channel or persisted session state are carved out of the stateless arm via arm_exclusions. Tests that are tied to one transport do not use the fixture: the wire-recording tests (their seam is the in-memory stream pair), the bare-ClientSession lifecycle tests, the real-clock timeout tests (the timeout machinery is transport-independent and must not race transport latency), and everything under transports/, which pins behaviour only observable on that transport.

A transport conformance test in transports/ speaks raw httpx2 against the mounted ASGI app only when its assertion is about HTTP semantics that Client cannot observe — status codes, response headers, SSE event fields, which stream a message travels on. Any other behaviour is asserted through a Client, connected to the mounted app via client_via_http(http) so several clients can share one session manager.

_requirements.py maps every behaviour the suite covers to the reason it must hold:

"tools:call:content:text": Requirement(
    source=f"{SPEC_BASE_URL}/server/tools#text-content",
    behavior="tools/call delivers arguments to the tool handler and returns its text content.",
),

• source is a deep link into the MCP specification for externally mandated behaviour, the literal string "sdk" for behaviour the SDK chose where the spec is silent, or "issue:#n" for a regression lock.
• behavior describes the required behaviour — what the specification (or the SDK's own contract) says should happen. Tests always pin the SDK's current behaviour; where that falls short of behavior, the gap is recorded as data rather than hidden in the test.
• divergence records that gap for entries whose tests pin the divergent current behaviour.
• deferred marks a behaviour that is tracked but has no test in this suite, with a precise reason: the SDK does not implement it, the negative cannot be observed, the assertion is schema-level rather than interaction-level, the feature is experimental (tasks), or the test would require real-time waits the suite refuses.
• transports names the transports a behaviour applies to; omitted means transport-independent.
• issue carries the tracking link for a recorded gap once one is filed.
• note carries free-form context that does not fit divergence or deferred.
• added_in / removed_in bound the spec versions the behaviour exists in, as a half-open [added_in, removed_in) window.
• supersedes / superseded_by link a retired entry to its replacement; the link is bidirectional and both ends must be versioned.
• arm_exclusions carve specific (transport, spec_version) matrix cells out with a typed ArmExclusionReason.
• known_failures mark specific (transport, spec_version) cells as strict xfail.

Tests link themselves to the manifest with a decorator:

@requirement("tools:call:content:text")
async def test_call_tool_returns_text_content() -> None: ...

test_coverage.py enforces the contract in both directions: every non-deferred requirement must be exercised by at least one test, every deferred requirement by none, and an unknown ID fails at import time. A behaviour without a manifest entry cannot be silently half-tested, and a manifest entry without a test cannot be silently aspirational.

1. A test reveals that the SDK does not do what the spec says. The test pins what the SDK actually does and a Divergence(note=..., issue=...) goes on the requirement.
2. When the behaviour is eventually fixed, the pinned test fails. Whoever makes the change finds the divergence note explaining that the old behaviour was a known gap, re-pins the test to the spec-correct output, and deletes the Divergence.
3. An empty divergence list means the SDK is spec-conformant on every behaviour the suite covers.

A requirement may carry both divergence and deferred: the divergence records that the SDK falls short of the spec, and the deferral records why no test pins it (typically because the divergent behaviour cannot be driven through the public API). Divergence alone implies a test pins the divergent behaviour; divergence plus deferred means the gap is known but unpinned.

This is also the triage key for any rewrite: a test that fails on the new code path either has a divergence note (the rewrite accidentally fixed a known gap — decide whether to keep the fix) or it does not (the rewrite broke something that was correct — fix the rewrite).

SPEC_VERSIONS in _requirements.py is the ordered tuple of protocol revisions the suite exercises. SPEC_BASE_URL (and SPEC_2026_BASE_URL) are pinned literals — not derived from SPEC_VERSIONS — so growing the active axis never repoints existing source links. The connect fixture fans out over CONNECTABLE_TRANSPORTS × SPEC_VERSIONS, but the grid is filtered per test: pytest_generate_tests reads the test's stacked @requirement marks and calls compute_cells(), which intersects the admissible cells across every cited requirement — a cell survives only if all of the test's requirements admit it.

streamable-http-stateless is the fourth connectable transport: the 2025-era unofficial stateless mode where each request opens a fresh transport, no session id is issued, and there is no standalone GET stream. Requirements that need a server→client back-channel or persisted session state are excluded from that arm via arm_exclusions (reasons server-initiated-request and requires-session).

What admits or excludes a cell:

• added_in / removed_in gate which spec versions a requirement exists in, as a half-open [added_in, removed_in) window. A test runs only on versions inside every cited requirement's window.
• arm_exclusions carve specific (transport, spec_version) cells out with a typed ArmExclusionReason. The reason vocabulary doubles as a re-admission checklist: when the gap closes, grep for the reason string to find every cell to re-admit.
• known_failures keep a cell in the grid but mark it as a strict xfail — the test runs and must fail; an unexpected pass fails the suite.
• TRANSPORT_SPEC_VERSIONS era-locks a transport to a subset of spec versions (currently only sse is locked to 2025-11-25). A (transport, version) cell is dropped if the version is not in the transport's entry; transports absent from the map serve every spec version. This is the mechanism for cutting an entire transport off from a new revision (or admitting it).
• transports is descriptive metadata for the non-connect transport-specific suites under transports/ and does not drive cell generation. Only arm_exclusions, added_in, removed_in, and TRANSPORT_SPEC_VERSIONS filter the grid.
• supersedes / superseded_by link a retired entry to its replacement. test_coverage.py enforces that links are bidirectional and versioned: the retired entry carries removed_in, the replacement carries added_in.

Node IDs stay [transport] while len(SPEC_VERSIONS) == 1, so today's test IDs are byte-identical to before the era axis existed. They become [transport-version] the moment a second version is appended to SPEC_VERSIONS.

When a new spec revision lands:

1. Append the version string to SPEC_VERSIONS (and to the SpecVersion Literal).
2. Walk the new revision's changelog.
3. For each affected requirement: set removed_in on retired behaviour, add a new entry with added_in for its replacement, and link the pair with supersedes / superseded_by. Behaviour that survives unchanged needs nothing beyond a re-audit of its source URL.
4. For requirements that cannot run on the new era's path, add an arm_exclusions entry with the appropriate ArmExclusionReason.
5. Review TRANSPORT_SPEC_VERSIONS: any era-locked transport will not produce cells on the new version unless its entry is extended (or removed); add an entry for any transport the new revision retires.

The shortest complete example of the conventions:

@requirement("tools:call:content:text")
async def test_call_tool_returns_text_content() -> None:
    """Arguments reach the tool handler; its content comes back as the call result."""

    async def call_tool(ctx: ServerRequestContext, params: types.CallToolRequestParams) -> CallToolResult:
        assert params.name == "add"
        assert params.arguments is not None
        return CallToolResult(content=[TextContent(text=str(params.arguments["a"] + params.arguments["b"]))])

    server = Server("adder", on_call_tool=call_tool)

    async with Client(server) as client:
        result = await client.call_tool("add", {"a": 2, "b": 3})

    assert result == snapshot(CallToolResult(content=[TextContent(text="5")]))

• The server is defined inside the test (or in a small fixture at the top of the file when several tests genuinely share it). The whole observable behaviour fits on one screen.
• Test names are behaviour sentences — they state the observable outcome, not the feature being poked. Docstrings add the one or two sentences of context a reviewer needs, including whether the assertion is spec-mandated, SDK-defined, or a known divergence.
• Handlers assert their dispatch identity first (assert params.name == "add"), proving the request that arrived is the request the test sent.
• The result proves the round trip. Server-side observations travel back to the test through the protocol itself (a tool returns what it saw) or through a closure-captured list; the test asserts after the call returns.
• Order within a test: server handlers → server construction → client callbacks → connect → act → assert. The test reads in the order the conversation happens.
• A registered handler or tool that a test never invokes gets a raise NotImplementedError body so it cannot silently become load-bearing.
• A test that needs a peer no real Server or Client can play (a server that answers initialize with an unsupported version, a client that sends malformed params) plays that side of the wire by hand over create_client_server_memory_streams(). This scripted-peer pattern is the suite's only way to drive behaviour the typed API cannot produce, and the docstring of every such test says so.

Stack a second @requirement decorator only when a test's natural assertions incidentally prove another behaviour — one capabilities snapshot proving four *:capability:declared entries, one input-schema identity check proving each preserved keyword. Do not build a test around covering many requirements at once; if the assertions would be separate, write separate tests.

The client's dispatcher starts a task per incoming notification in arrival order but does not await it before reading the next message, so completion order is not structural. What still holds: the in-memory transport delivers everything on one ordered stream, and a callback that records synchronously (no await before the append) finishes its scheduling slice before the awaited request's waiter — woken strictly later — resumes. So tests whose callbacks are plain appends may still collect into a list and assert after the call. A callback that awaits before recording loses that ordering and must synchronise. The other exceptions:

• a notification not triggered by a request the test is awaiting needs an anyio.Event set in the receiving handler and awaited under anyio.fail_after(5);
• the ordering guarantee does not survive transports that split messages across streams (the streamable HTTP standalone GET stream) — see transports/test_streamable_http.py.

CI requires 100% line and branch coverage, including tests/, and strict-no-cover fails the build if a line marked # pragma: no cover is ever executed. When a new test starts covering a pragma'd line in src/, delete the pragma in the same change. Do not add new # type: ignore or # noqa comments; restructure instead. Two pragmas are sanctioned in this suite's test code, both for known-upstream tracer bugs and only after restructuring has been tried: # pragma: no branch on a with/async with line whose only fault is coverage.py mis-tracing the exit arc of a nested async context (reserve it for shapes that cannot collapse — a sync with adjacent to an async with); and # pragma: lax no cover on a single statement that 3.11's tracer drops because the preceding async with unwinds via coro.throw() (python/cpython#106749, wontfix on 3.11) — this hits any test that must run statements after a ClientSession/streamable_http_client exits but still inside an outer async with, and no restructure can avoid it.

A handful of # pragma: lax no cover markers in src/ cover teardown exception handlers whose execution is timing-dependent under the in-process HTTP bridge — the POST-stream and stateless-session except Exception handlers in server/streamable_http*.py and the _terminated check in message_router. strict-no-cover does not check lax lines; do not promote them to strict no cover without first making the teardown ordering deterministic. The suite also relies on a one-line src/mcp/server/sse.py fix (sse_stream_reader.aclose()) that closes a stream the SSE leg would otherwise leak.