fix(e2e): stabilize flaky provider suite ordering and zai requestCapture race#45
Merged
Merged
Conversation
Contributor
Author
|
2 issues outstanding. Action required. See task
|
Codecov Report✅ All modified and coverable lines are covered by tests. 📢 Thoughts on this report? Let us know! |
3c0a972 to
2b1c2bc
Compare
edelauna
approved these changes
Jun 7, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What problem this solves
Fixes two intermittent e2e flakes observed together in CI:
codemode pinned to the Bedrock provider when Mocha's non-deterministic glob happened to loadproviders/bedrock.test.tsbefore the tool suites. Tool suite tasks started incodemode then hit the Bedrock endpoint, got404 No fixture matched, and timed out after 60–90 s each.requestCaptureobject across tests. A straggling async API call from the glm-5.1 test could overwriterequestCapture.maxTokensafter the glm-5-turbo test'ssetup()reset it but before its own request landed, causing the40_551assertion to read40000instead.Why this change was made
Both failures were timing/ordering-dependent — they passed on re-run with a different file order or timing. Rather than rely on luck, the fixes make the suite order deterministic and the assertion immune to async stragglers.
Changes
apps/vscode-e2e/src/suite/index.ts— sort test files soproviders/suites always run after all other suites (alphabetical within each bucket). Provider teardown state can no longer bleed into tool suites regardless of glob order.apps/vscode-e2e/src/suite/providers/zai.test.ts— snapshotrequestCapture.maxTokensimmediately afterwaitUntilCompletedreturns in both glm-5.1 and glm-5-turbo tests, before any straggling async calls can overwrite it.User impact
No production code changed. CI
e2e-mockjob should no longer flake on these two failure modes.Closes #512
Closes #514