Skip to content

refactor(vgi-bench): address review feedback from subtree sync - #33

Merged
abhinav-pola merged 1 commit into
mainfrom
fix/vgi-bench-review-cleanup
Aug 14, 2026
Merged

refactor(vgi-bench): address review feedback from subtree sync#33
abhinav-pola merged 1 commit into
mainfrom
fix/vgi-bench-review-cleanup

Conversation

@abhinav-pola

@abhinav-pola abhinav-pola commented Aug 14, 2026

Copy link
Copy Markdown
Contributor

Summary

Addresses the review feedback from the vendored subtree sync PR (openrouter-web#34180). Since packages/bench-harness is a vendored subtree, these fixes must land upstream here and will propagate on the next sync.

Changes

Un-export unused symbols (fallow audit):

  • VGI_BENCH_DATASET_PATH, VGI_BENCH_TEMPERATURE, makeVgiBenchDatasetLayer, and vgiBenchSolver were exported but never imported by other modules. Removed the export keyword.

Fix test name (Devin review #1):

  • Renamed "builds the pinned multiple-choice prompt with lowercase letters" → the snapshot contains uppercase letters (A) B) C) D)).

Remove duplicate test (Devin review #2):

  • "does not splice into dots in the hostname for extension-less paths" used the exact same input and expected output as "appends the suffix when the filename has no extension". Deleted the duplicate.

Merge fragmented tests (Devin review #3):

  • Combined five separate tests that each asserted one property of vgiBenchRecordToSample(VGI_RECORD, 0) into a single test with an inline snapshot of the full sample, plus canary-absence assertions.

Validation

format:check  OK
check (oxlint) OK
typecheck      OK
build          OK
bun test       1250 pass

Open in Devin Review

- Un-export VGI_BENCH_DATASET_PATH, VGI_BENCH_TEMPERATURE,
  makeVgiBenchDatasetLayer, and vgiBenchSolver (never imported externally)
- Fix test name: 'lowercase letters' -> the snapshot has uppercase letters
- Remove duplicate downscaledVideoUrl test (identical input/output)
- Merge five fragmented vgiBenchRecordToSample tests into a single test
  with an inline snapshot of the full sample plus canary-absence checks
@abhinav-pola
abhinav-pola requested a review from a team as a code owner August 14, 2026 22:09

@devin-ai-integration devin-ai-integration Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

✅ Devin Review: No Issues Found

Devin Review analyzed this PR and found no bugs or issues to report.

Open in Devin Review

@abhinav-pola
abhinav-pola merged commit fc216df into main Aug 14, 2026
5 checks passed
@abhinav-pola
abhinav-pola deleted the fix/vgi-bench-review-cleanup branch August 14, 2026 22:13
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants