test(e2e): drop xfail markers for now-fixed team-budget-JSON and custom-pricing-leak bugs

[mubashir1osmani]

Relevant issues

Two e2e tests were xfail(strict=True) for known proxy bugs that are now fixed on litellm_internal_staging:

• /team/new wrote budget_limits as a raw list into the Json? column, which Prisma rejected with a 500 (the key path and /team/update json.dumps first). Related: feat: multiple concurrent budget windows per API key and team (#24883) #25109
• a deployment's custom per-token pricing leaked into the shared cost map for sibling deployments of the same underlying model

Pre-Submission checklist

• I have added meaningful tests
• My PR's scope is as isolated as possible; it only solves 1 specific problem

Screenshots / Proof of Fix

Both fixes already shipped to litellm_internal_staging, so the two tests pass their assertions there. Because they were still marked xfail(strict=True), the run reported the unexpected passes as failures:

[XPASS(strict)] known proxy bug: /team/new writes budget_limits straight to the Json? column ...
[XPASS(strict)] known proxy bug: a deployment's custom per-token pricing leaks into the shared cost map ...
FAILED budgets/test_team_multi_window_budget_e2e.py::test_team_short_window_blocks_then_resets
FAILED llm_translation/test_custom_pricing_e2e.py::test_custom_pricing_is_isolated_from_sibling_deployment
=================== 2 failed, 59 passed in 497.24s ===================

With the strict markers removed, both run as ordinary passing tests, taking the suite to all-green; their assertions are unchanged, so they now stand as regression guards for the two fixes

Type

🧹 Refactoring
✅ Test

Changes

Removed the @pytest.mark.xfail(strict=True, ...) decorator from test_team_short_window_blocks_then_resets and test_custom_pricing_is_isolated_from_sibling_deployment, as each marker's reason instructed once the underlying bug was fixed. Updated both docstrings to describe the regression each test now pins rather than the bug it used to tolerate. No assertion changed

[greptile-apps]

Greptile Summary

This PR promotes two previously-skipped e2e tests from xfail(strict=True) to ordinary passing tests now that the underlying proxy bugs have been fixed on litellm_internal_staging. No assertions or test logic were changed — only the decorator and docstrings.

• Removes the xfail marker from test_team_short_window_blocks_then_resets, which now acts as a regression guard ensuring /team/new correctly json.dumps the budget window list before writing to the Json? column.
• Removes the xfail marker from test_custom_pricing_is_isolated_from_sibling_deployment, which now guards against custom per-token pricing leaking into the shared cost map for sibling deployments of the same underlying model.

Confidence Score: 5/5

Safe to merge — only xfail markers and docstrings are changed, no test logic or assertions are touched.

Both tests retain all their original assertions and now run unconditionally as regression guards. The PR description references the concrete XPASS output that triggered this change, satisfying the team's evidence-of-fix rule. No production code is modified.

No files require special attention.

Important Files Changed

Filename	Overview
tests/e2e/budgets/test_team_multi_window_budget_e2e.py	Removes xfail(strict=True) marker from test_team_short_window_blocks_then_resets and updates the module docstring to describe it as a regression guard for the now-fixed /team/new json.dumps bug; no assertions changed.
tests/e2e/llm_translation/test_custom_pricing_e2e.py	Removes xfail(strict=True) marker from test_custom_pricing_is_isolated_from_sibling_deployment and updates the module docstring to describe it as a regression guard for the now-fixed pricing isolation bug; no assertions changed.

_{Reviews (1): Last reviewed commit: "test(e2e): drop xfail markers for the no..." | Re-trigger Greptile}

[mubashir1osmani]

these two tests were internally found bugs that are now fixed on litellm_internal_staging - removing xfail markers

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!