Repository navigation
fix(test): reconcile #11960's tests with the one-store memory refactor - #12098
Closed
bolichen97 wants to merge 1 commit into
Closed
bolichen97 wants to merge 1 commit into
bolichen97 wants to merge 1 commit into
Conversation
#11960 (6c21f0d) was authored against a base older than the member memory refactor bc0d2b0 and merged after it with a textually clean but semantically stale test tree, so every PR on main now fails three required lanes: - Backend shard 1: test_subagent_continuable.py carried #11960's copy of the file, which dropped the refactor's adjustments and monkeypatched kiro_crew.member_memory_auth.private_memory_execution_supported, a function the refactor removed (60 errors, 34 failures). - Backend shard 6 and Real Adapter Contract Tests: the codex real adapter tests lost their @pytest.mark.real_adapter markers and the require_real_adapter guard, so the lane selected 1 of the 4 measured contracts and test_real_adapter_gate.py failed. Three-way merge of the test file (base bc0d2b0^, refactor side, #11960 side): the refactor's version plus #11960's appended TestSharedBindIdentityLabel, whose MagicMock handle now carries a real privacy mode. Codex tests get their markers and the _require_codex_acp wrapper back; the token-spending opt-in test stays unmarked with an inline skip so the lane's no-skip assertion holds.
5 tasks done
Contributor
Author
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why
#11960 (
6c21f0d4d9) was authored against a base older than the member memory refactorbc0d2b0cfcand merged after it. The merge was textually clean but the test tree it carried is semantically stale, so every PR onmainnow fails three required lanes (reproduced on the pure revert #12091 and on #11996, both red on exactly these):Backend Tests (3.12, 1)and the Windows twin:test/test_subagent_continuable.pyis feat(acp): continue a codex subagent from its persisted thread #11960's copy of the file, which dropped the refactor's adjustments and monkeypatcheskiro_crew.member_memory_auth.private_memory_execution_supported, a function the refactor removed. 60 errors and 34 failures;6c21f0d4d9^passes all 167.Backend Tests (3.12, 6):test/test_real_adapter_gate.py::test_the_tree_carries_the_guarded_tests_this_lane_exists_forfails because the codex real adapter tests lost their@pytest.mark.real_adaptermarkers and therequire_real_adapterguard.Real Adapter Contract Tests: the lane selects on that marker and now finds 1 of the measured floor of 4.What changed
test/test_subagent_continuable.py: three-way merge (basebc0d2b0cfc^, refactor side6c21f0d4d9^, feat(acp): continue a codex subagent from its persisted thread #11960 side6c21f0d4d9). Result is the refactor's version plus feat(acp): continue a codex subagent from its persisted thread #11960's appendedTestSharedBindIdentityLabel, whoseMagicMockhandle now carriesmemory_mode = "persistent"because the one-store bind validates the privacy mode it tightens against. Both dead monkeypatches are gone.test/test_codex_session_mcp.py: restores thereal_adapter_gateimport, the_require_codex_acpwrapper, the@pytest.mark.real_adaptermarker and the in-body guard on the three measured codex contracts, replacing theskipif(_ENTRY is None ...)decorators feat(acp): continue a codex subagent from its persisted thread #11960 introduced. The token-spending opt-in testtest_real_codex_acp_load_after_close_restoresstays unmarked (the lane asserts nothing it selected skipped) and guards the adapter with an inlinepytest.skip.No production code is touched.
Tests
test_subagent_continuable.py,test_real_adapter_gate.py,test_codex_session_mcp.py,test_opencode_session_mcp.py: 315 passed, 5 skipped locally.pytest -m real_adapter --collect-only test/test_codex_session_mcp.py test/test_opencode_session_mcp.py: 4 selected (the lane's floor).Pattern harvest
Rule candidate: CI
Pattern: a PR whose CI ran against an older base merges textually clean over a refactor that changed the API its tests rely on. The
mainpush pipeline runs no backend shards, so the breakage stays green until the next PR. Second instance this week (#11947 was the first, reverted in #12091). Candidate rule: require the PR's checks to have run against the current base before merge (amergeable_statefreshness check), or an import and collection smoke onpushtomain.