Skip to content

fix(admin): support Luna key tests (v0.30.62) - #1013

Merged
mylukin merged 1 commit into
mainfrom
codex/api-key-test-reasoning
Oct 29, 2026
Merged

mylukin merged 1 commit into
mainfrom
codex/api-key-test-reasoning

Conversation

@mylukin

@mylukin mylukin commented Oct 29, 2026

Copy link
Copy Markdown
Contributor

The per-key test button's default GPT-6 Luna request returned HTTP 400 because the model rejects reasoning_effort: minimal. Use the supported none mode for this short connectivity request, keeping the 256-token cap and all account isolation behavior.

Live verification with the same credential reproduced unsupported_value, then returned HTTP 200 and OK with none. Added a failing-then-passing regression test; all 7 focused route tests pass. pnpm release:prepare 0.30.62 refreshes the release contract; official stable snapshots are unchanged.

This corrects the live acceptance finding from #1012. Deploy both remote and LAN after CI and immutable image publication, then verify all four keys through the actual UI.

Co-Authored-By: Codex <noreply@openai.com>
@mylukin
mylukin merged commit c01dc55 into main Oct 29, 2026
10 checks passed
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.

1 participant