fix(router): use PEP 585 frozenset[str] to satisfy UP006 strict budget

[cursor]

Relevant issues

Linear ticket

Pre-Submission checklist

• I have added meaningful tests
• My PR passes all CI/CD checks (e.g., lint, format, unit tests)
• My PR's scope is as isolated as possible; it only solves 1 specific problem
• I have received a Greptile Confidence Score of at least 4/5

Screenshots / Proof of Fix

Before this fix, LiteLLM Linting > Check strict-rule budget (delta vs base) on the promote-staging-to-main PR #33491 (head 39c01fe) failed with:

FAIL: strict-rule totals exceed their limit (base 0fbdea4bf5be35561a3cc0c12a431f8f221618b1):
  UP006: total 12794 over limit 12792 (this change added 1)
    litellm/router.py:499

Run log: https://github.com/BerriAI/litellm/actions/runs/29464302114

After this fix, running the same gate locally against the same base:

$ python3 scripts/ruff_strict_gate.py --base 0fbdea4bf5be35561a3cc0c12a431f8f221618b1
OK: every strict rule is within its codebase ceiling (base 0fbdea4bf5be35561a3cc0c12a431f8f221618b1)

Type

🐛 Bug Fix

Changes

PR #32926 introduced self.team_public_model_names: FrozenSet[str] = frozenset() at litellm/router.py:499 using the deprecated typing.FrozenSet, which trips ruff's UP006 non-pep585-annotation rule and pushed the repo one violation over the strict-rule ceiling (12794 vs. limit 12792). Switching the annotation to the PEP 585 built-in generic frozenset[str] and dropping the now-unused FrozenSet import brings the count back down to the base level and clears the gate.

Final Attestation

• The tests check the right things, including the edge cases, and regressions in the respective real-world customer use-cases are not possible after this PR

  

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[codspeed-hq]

Merging this PR will not alter performance

✅ 31 untouched benchmarks

---

_{Comparing cursor/ci-autofix-automation-23a1 (b2a7a83) with litellm_internal_staging (39c01fe)}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!