[kalvenschraut]

What changed

When CLIProxyAPI's usage response omits plan_type, T3 can lose the account's plan label and classify a Free or Go quota window as a session window even though the auth-file metadata contains the plan.

Resolve the plan once from the usage response, then id_token.plan_type, then the legacy id_token.chatgpt_plan_type. Use that value for both the label and quota conversion. The current CLIProxyAPI management handler emits plan_type.

UI

Synthetic browser fixture generated from the baseline and patched adapters. The usage response omits the plan and window duration; auth metadata contains plan_type: "free". The quota is now classified as Monthly.

Before	After

Verification

• 32 adapter/converter tests passed. Regression coverage checks Free and Go with both current and legacy token fields, asserting the plan labels and monthly window duration when the usage plan is absent.
• Server package typecheck and changed-file lint passed.
• The adapter produces the shared report for all clients. Browser evidence uses synthetic fixtures; no live hub or mobile end-to-end test was run.
• Fable approved the final diff after checking the management-handler schema evidence.

Implemented with GPT-6 in Codex; reviewed with claude-fable-5-1 through Claude Code.

Summary by CodeRabbit

• Bug Fixes
  • Improved Codex account plan detection when upstream usage data omits subscription metadata.
  • Subscription labels now correctly identify Free and Go plans using available current or legacy account information.
  • Rate-limit snapshots and monthly quota windows now reflect the resolved account plan more reliably.

[macroscopeapp]

Approvability

Verdict: Approved at 836db7a

Macroscope's review found this PR approvable — This is a focused usage bug fix that adds a backward-compatible metadata fallback and verifies both current and legacy Codex plan fields. Existing upstream plan data remains unchanged, while missing-plan accounts now receive the correct label and monthly quota classification.

Notes:

• Diff unchanged. Approvability was decided on eligibility alone.

^{You can add or adjust custom eligibility rules. Learn more.}

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Advanced

Run ID: 02be3389-14e4-47be-a60a-f75749ae584d

📥 Commits

Reviewing files that changed from the base of the PR and between 08463e2 and 836db7a.

📒 Files selected for processing (2)

• apps/server/src/usage/cliproxyApi.test.ts
• apps/server/src/usage/cliproxyApi.ts

Included review availability: Your plan provides up to 10 included reviews per hour; 6 remain after this review.

---

📝 Walkthrough

Walkthrough

Codex plan resolution now falls back to token metadata when upstream usage responses omit plan information. Tests cover current and legacy plan fields for Free and Go accounts.

Changes

Codex plan resolution

Layer / File(s)	Summary
Token metadata and plan mapping apps/server/src/usage/cliproxyApi.ts, apps/server/src/usage/cliproxyApi.test.ts	The auth schema accepts id_token.plan_type. Plan resolution falls back to current and legacy token fields. Tests verify subscription labels and monthly quota windows for Free and Go accounts.

Priority: ⬇️ Low

Estimated code review effort: 2 (Simple) | ~10 minutes

Merge Risk: ⚪ Minimal · up to 836db

Codex usage now recovers missing plan metadata from current and legacy token fields, preserving correct account labels and monthly quota windows for Free and Go plans. No current merge-blocking risk remains.

Suggested reviewers: juliusmarminge

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 1 functions across 2 files.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly identifies the main change: recovering missing Codex plan data from authentication metadata.
Description check	✅ Passed	The description explains what changed, why it changed, UI evidence, verification results, and testing limitations. It does not use the exact template headings for “Why,” “UI Changes,” and “Checklist,”…
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

• Fix all pre-merge checks with AI

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

_{Comment @coderabbitai help to get the list of available commands.}

Dismissing prior approval to re-evaluate 836db7a

[maria-rcks]

Note

Written by claude-opus-5-5 on behalf of Maria

Hi! We are cleaning up open PRs, and this one appears to have been created with an older model (gpt-6). If this change is really important, we recommend rebuilding the PR with a newer model if possible.

[maria-rcks]

Note

Written by claude-opus-5-5 on behalf of Maria

Reopening, this was closed by mistake. Sorry for the noise!

[juliusmarminge]

Taking over this one. I’ll reduce the tests to the useful coverage, update the implementation against current main, and verify it before publishing the changes.

[juliusmarminge]

Thanks, @RTVision, for identifying the missing account-plan fallback and contributing the fix. The maintainer takeover is now published as #21413.

GitHub rejected the push to this fork with HTTP 403 (Permission to RTVision/t3code.git denied to juliusmarminge), despite reporting maintainer edits enabled. The replacement uses an origin branch so the work can proceed without changing your fork's access.

I compared the original intent with the full replacement diff: it preserves usage-response → current token → legacy token plan precedence and feeds the resolved value to both the label and quota conversion. The original commits remain in the ancestry, and the replacement points to the exact reviewed commit, with no new source changes or rebase. The test was reduced to one three-account precedence regression; 39 focused tests and two independent reviews passed. Verification is mocked-HTTP service output, not a live hub or browser run.

Closing this PR in favor of #21413; the replacement remains open and unmerged.