revert: "feat(proxy): server-side Team Usage export beyond the top-N key cap (#42996)"

[yassin-berriai]

Reverts #42996

TLDR

Problem this solves:

• feat(proxy): server-side Team Usage export beyond the top-N key cap #42996 (server-side Team Usage export) needs to come back out of main

How it solves it:

• Plain git revert of the squash commit 77eccac
• Removes the GET /team/daily/activity/export route, its types, tests and the dashboard wiring
• Keeps the later CLI session test that fix(spend): attribute CLI session spend to the per-user cli-session alias instead of the hashed session token #40541 added to the same test file

User Flow

Before: with more keys than USAGE_TOP_API_KEYS_LIMIT, a proxy admin can export Team Usage from the server

1. They open http://localhost:4000/ui/?page=team-usage and click Export Data
2. The dashboard calls GET http://localhost:4000/team/daily/activity/export?...&export_type=daily_with_keys&format=csv and gets a 200 with a CSV

After: the proxy behaves as it did before #42996

1. They open http://localhost:4000/ui/?page=team-usage; Export Data is disabled when the key list was truncated, with the tooltip pointing at USAGE_TOP_API_KEYS_LIMIT
2. GET http://localhost:4000/team/daily/activity/export returns 404

Type

🧹 Refactoring

Caveats (if any)

Low

• First of a three PR revert stack: feat(usage): search team keys beyond the top-N in the Team usage view #42857 and feat(usage): search keys beyond the top-N usage subset #42827 revert on top of this one because their code overlaps in the same files
• The test file test_common_daily_activity.py had one conflict; resolved by taking the pre feat(proxy): server-side Team Usage export beyond the top-N key cap #42996 file and re-adding the fix(spend): attribute CLI session spend to the per-user cli-session alias instead of the hashed session token #40541 test

🤖 Generated with Claude Code

[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.}

[greptile-apps]

Confidence Score: 4/5

[Medium risk] Removes server-side team usage export feature and related code.

The PR should not merge until the route removal satisfies the repository’s compatibility requirement and the independent test coverage is retained.

Findings

1.  Existing export requests break ▶

Summary

This PR reverts server-side Team Usage export, removing the proxy route, its types and tests, and the dashboard’s server-download path. The dashboard again blocks export when its key list is truncated.

• The route removal conflicts with the repository’s compatibility requirement.
• Deleting the export test file also discards an independent spend-log regression test.

_{Reviews (1) · Last reviewed commit: "Revert "feat(proxy): server-side Team Us..."}

[greptile-apps]

Existing export requests break

Clients that already call /team/daily/activity/export will receive 404 after this route is removed. The repository requires backwards-incompatible changes to have a user-controlled flag, but this removal provides no way to retain the endpoint.

Rule Used: What: avoid backwards-incompatible changes without user-controlled flags Why: This breaks current behaviour for users using existing functionality Example of BAD: this PR (#22164) introduced run_post_custom... (source)

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!

[greptile-apps]

Comments Outside Diff

These findings sit on lines the diff does not cover, so they could not be posted inline. Each one leaves this list once its file changes.

•  Independent spend test deleted tests/integration/spend/test_team_daily_activity_export.py:434 ▶

  Deleting this file also removes a test that checks whether a non-streaming OpenAI SDK chat produces exactly one spend-log row. That behavior is independent of the export endpoint, so the deletion weakens regression coverage. The repository requires test changes not to weaken coverage; retain this check in an appropriate test file before merging.

  Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!

[codspeed]

Merging this PR will not alter performance

✅ 31 untouched benchmarks

---

_{Comparing litellm_revert_42996 (72fc7e0) with main (7244040)}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!