ci(conformance): add 2026-07-28 carried-forward leg + bump harness to 0.2.0-alpha.4 (#2911)

This commit is contained in:
Max
2026-06-19 14:48:21 +01:00
committed by GitHub
parent 364b762ef6
commit 9e6d00318c
4 changed files with 179 additions and 4 deletions
+11
View File
@@ -6,6 +6,9 @@ It handles all conformance test scenarios via environment variables and CLI argu
Contract:
- MCP_CONFORMANCE_SCENARIO env var -> scenario name
- MCP_CONFORMANCE_CONTEXT env var -> optional JSON (for client-credentials scenarios)
- MCP_CONFORMANCE_PROTOCOL_VERSION env var -> spec version the harness mock
server is speaking (e.g. "2025-11-25", "2026-07-28"). Always set; defaults
to the harness's LATEST_SPEC_VERSION when --spec-version is omitted.
- Server URL as last CLI argument (sys.argv[1])
- Must exit 0 within 30 seconds
@@ -50,6 +53,13 @@ logging.basicConfig(
)
logger = logging.getLogger(__name__)
#: Spec version the harness is running this scenario at (e.g. "2025-11-25",
#: "2026-07-28"). The harness always sets this (it falls back to its own
#: LATEST_SPEC_VERSION when --spec-version is omitted), so None means we were
#: invoked outside the harness. Handlers that need to take the stateless 2026
#: path will branch on this once the SDK has one; today it is logged only.
PROTOCOL_VERSION: str | None = os.environ.get("MCP_CONFORMANCE_PROTOCOL_VERSION")
# Type for async scenario handler functions
ScenarioHandler = Callable[[str], Coroutine[Any, None, None]]
@@ -347,6 +357,7 @@ def main() -> None:
server_url = sys.argv[1]
scenario = os.environ.get("MCP_CONFORMANCE_SCENARIO")
logger.debug(f"Conformance protocol version: {PROTOCOL_VERSION!r}")
if scenario:
logger.debug(f"Running explicit scenario '{scenario}' against {server_url}")
@@ -0,0 +1,149 @@
# Expected failures for the carried-forward 2026-07-28 legs
# (`--suite all --spec-version 2026-07-28` for both server and client).
#
# This baseline is separate from expected-failures.yml because entries are
# keyed by scenario name only: a scenario that passes at its default version
# in the 2025 legs but fails when forced to 2026-07-28 (or vice versa) cannot
# be expressed in a shared file (the passing leg would flag the entry as
# stale). Like expected-failures.yml, this single file covers both
# directions: the client 2026 leg reads the `client:` section and the server
# 2026 leg reads the `server:` section. Both burn down independently of the
# 2025 legs.
#
# Baseline established against @modelcontextprotocol/conformance pinned in
# .github/workflows/conformance.yml (CONFORMANCE_VERSION = 0.2.0-alpha.4).
# New conformance releases are adopted by deliberately bumping that pin and
# reconciling both this file and expected-failures.yml in the same change.
#
# Entries are grouped by what unblocks them. As each gap closes the
# corresponding scenarios start passing and MUST be removed from this list
# (the runner fails on stale entries), so the baseline burns down per
# milestone.
client:
# --- No stateless client path on main yet ---
# client.py drives the 2025 stateful lifecycle (initialize handshake +
# session). The 2026-mode mock server is stateless, so the call sequence
# never reaches the assertion. Unblocks when client.py's is_modern_protocol()
# branch takes the per-request _meta path.
- tools_call
# --- SEP-837 (application_type during DCR) ---
# The sep-837-application-type-present check only fires on 2026-version
# runs; the client omits application_type during Dynamic Client
# Registration, so every auth scenario that reaches DCR fails it on this
# leg (the same scenarios pass at their default version in the 2025 legs).
- auth/metadata-default
- auth/metadata-var1
- auth/metadata-var2
- auth/metadata-var3
- auth/scope-from-www-authenticate
- auth/scope-from-scopes-supported
- auth/scope-omitted-when-undefined
- auth/token-endpoint-auth-basic
- auth/token-endpoint-auth-post
- auth/token-endpoint-auth-none
- auth/offline-access-not-supported
# --- Auth scenarios cut short by the 2026 connection lifecycle ---
# The auth fixture flow drives the 2025 stateful lifecycle; the 2026-mode
# mock rejects the MCP POST before the scope-escalation behaviour these
# scenarios measure, so no authorization requests are observed. Unblocks
# when client.py's auth flow speaks the 2026 per-request lifecycle.
- auth/scope-step-up
- auth/scope-retry-limit
# --- Same gaps as the 2025 baseline (fail identically when forced to 2026-07-28) ---
# SEP-2575 (request metadata / _meta envelope): client does not populate the
# _meta envelope or the MCP-Protocol-Version header semantics yet.
- request-metadata
# SEP-2322 (multi-round-trip requests): client does not echo requestState /
# handle IncompleteResult yet.
- sep-2322-client-request-state
# SEP-2243 (HTTP standardization): no fixture handler / client header support yet.
- http-custom-headers
- http-invalid-tool-headers
# SEP-2106 (JSON Schema $ref handling): client still dereferences network $refs.
- json-schema-ref-no-deref
# SEP-2468 (authorization response iss parameter): not implemented in the client.
- auth/iss-supported
- auth/iss-not-advertised
- auth/iss-supported-missing
- auth/iss-wrong-issuer
- auth/iss-unexpected
- auth/iss-normalized
- auth/metadata-issuer-mismatch
# SEP-2352 (authorization server migration): client does not re-register when
# PRM authorization_servers changes.
- auth/authorization-server-migration
# auth/enterprise-managed-authorization (SEP-990) is in the 2025 baseline but
# NOT here: the harness skips it as inapplicable at --spec-version 2026-07-28
# (it is an extension scenario not carried into the 2026 wire), so it is
# neither run nor evaluated on this leg.
server:
# --- No stateless server path on main yet (carried-forward 2025-era scenarios) ---
# mcp-everything-server only runs in 2025 stateful mode. With
# --spec-version 2026-07-28 the harness sends stateless requests
# (MCP-Protocol-Version: 2026-07-28, _meta envelope, no initialize), which
# the server rejects before the handler runs. These scenarios all pass on
# the 2025 legs; they unblock once mcp-everything-server routes 2026
# requests through a stateless path.
- completion-complete
- tools-list
- tools-call-simple-text
- tools-call-image
- tools-call-audio
- tools-call-embedded-resource
- tools-call-mixed-content
- tools-call-error
- tools-call-with-progress
- server-sse-multiple-streams
- resources-list
- resources-read-text
- resources-read-binary
- resources-templates-read
- prompts-list
- prompts-get-simple
- prompts-get-with-args
- prompts-get-embedded-resource
- prompts-get-with-image
- dns-rebinding-protection
# SEP-2106 (JSON Schema 2020-12 in tool inputSchema): the fixture tool's
# schema has none of the 2020-12 keywords the scenario checks. The scenario
# is in `--suite all` but not `--suite active`, so this is the only leg that
# runs it; it fails identically at 2025-11-25 (not a 2026-path regression).
- json-schema-2020-12
# --- Draft scenarios (same failures and reasons as the `--suite draft` leg) ---
# SEP-2575 (stateless HTTP / _meta envelope): server has no stateless mode,
# _meta-derived capabilities, error-code mappings, or server/discover yet.
- server-stateless
# SEP-2322 (multi-round-trip requests / IncompleteResult): not implemented.
- input-required-result-basic-elicitation
- input-required-result-basic-sampling
- input-required-result-basic-list-roots
- input-required-result-request-state
- input-required-result-multiple-input-requests
- input-required-result-multi-round
- input-required-result-non-tool-request
- input-required-result-result-type
- input-required-result-tampered-state
- input-required-result-capability-check
# SEP-2549 (caching): no ttlMs/cacheScope support.
- caching
# SEP-2243 (HTTP header standardization): -32001 HeaderMismatch handling and
# case-insensitive/whitespace-trimmed header validation not implemented.
- http-header-validation
- http-custom-header-server-validation
# --- WARNING-only entries ---
# These scenarios emit no FAILURE checks, only SHOULD-level WARNINGs, but
# the expected-failures evaluator counts WARNINGs as failures. Same entries
# as the draft suite in expected-failures.yml.
# SEP-2164: server returns -32600 (not -32602) and omits error.data.uri.
- sep-2164-resource-not-found
# SEP-2322 SHOULD-level behaviours (re-request missing inputResponses,
# ignore unrecognized inputResponses keys).
- input-required-result-missing-input-response
- input-required-result-ignore-extra-params
@@ -2,9 +2,10 @@
# CI exits 0 if only these fail, exits 1 on unexpected failures or stale entries.
#
# Baseline established against @modelcontextprotocol/conformance pinned in
# .github/workflows/conformance.yml (CONFORMANCE_VERSION = 0.2.0-alpha.3).
# .github/workflows/conformance.yml (CONFORMANCE_VERSION = 0.2.0-alpha.4).
# New conformance releases are adopted by deliberately bumping that pin and
# reconciling this file in the same change.
# reconciling both this file and expected-failures.2026-07-28.yml in the same
# change.
#
# Entries are grouped by SEP. As each SEP lands in the SDK the corresponding
# scenarios start passing and MUST be removed from this list (the runner fails
+16 -2
View File
@@ -15,8 +15,9 @@ permissions:
env:
# Pinned conformance harness version. Bump deliberately and reconcile
# .github/actions/conformance/expected-failures.yml in the same change.
CONFORMANCE_VERSION: "0.2.0-alpha.3"
# both .github/actions/conformance/expected-failures*.yml files in the
# same change.
CONFORMANCE_VERSION: "0.2.0-alpha.4"
jobs:
server-conformance:
@@ -43,6 +44,12 @@ jobs:
./.github/actions/conformance/run-server.sh
--suite draft
--expected-failures ./.github/actions/conformance/expected-failures.yml
- name: Run server conformance (2026-07-28 wire, all suite)
run: >-
./.github/actions/conformance/run-server.sh
--suite all
--spec-version 2026-07-28
--expected-failures ./.github/actions/conformance/expected-failures.2026-07-28.yml
client-conformance:
runs-on: ubuntu-latest
@@ -64,3 +71,10 @@ jobs:
--command 'uv run --frozen python .github/actions/conformance/client.py'
--suite all
--expected-failures ./.github/actions/conformance/expected-failures.yml
- name: Run client conformance (2026-07-28 wire, all suite)
run: >-
npx --yes @modelcontextprotocol/conformance@"$CONFORMANCE_VERSION" client
--command 'uv run --frozen python .github/actions/conformance/client.py'
--suite all
--spec-version 2026-07-28
--expected-failures ./.github/actions/conformance/expected-failures.2026-07-28.yml