Port upstream 0.69.0: honor explicit Antigravity quota window cadence

[Finesssee]

Summary

Antigravity quota-summary buckets now accept an explicit window field. A non-empty window replaces the bucketId/displayName cadence candidates entirely: weekly -> 10080 min, 5h/session aliases -> 300 min, anything else -> no cadence. A blank window falls back to the previous id/name inference. Weekly-only Starter groups stay visible without an invented 5-hour bucket (the existing informational "No active 5h session" primary placeholder is unchanged). The 0.64.1 invariant that the selected primary/secondary are removed from extra_rate_windows is preserved.

Upstream reference

• Release 0.69.0 bullet: "Antigravity: preserve grouped quotas incl. weekly-only Starter allowances; honor explicit quota-window cadence" (fix(antigravity): preserve grouped OAuth quota summaries steipete/CodexBar#4084, refs Antigravity OAuth path shows fewer model-family groups than agy CLI path; "Plan usage" empty (0.45.2) steipete/CodexBar#2427, fix Antigravity if free tier usage (Antigravity Starter Quota) steipete/CodexBar#3789).
• Tag-pinned (v0.69.0): Sources/CodexBarCore/Providers/Antigravity/AntigravityStatusProbe.swift (quotaCadenceCandidates, quotaBucketKind), AntigravityQuotaSummaryParser.swift (window: String?), Tests/CodexBarTests/AntigravityQuotaSourceParityTests.swift.

Ported / Deferred

• Ported: window field on QuotaSummaryBucket and bucket_kind precedence in rust/src/providers/antigravity/quota_summary.rs.
• Deferred/skipped: the OAuth retrieveUserQuotaSummary half (Windows has no Antigravity OAuth lane).

Validation

• cargo +1.98.0 fmt --all: clean
• cargo +1.98.0 clippy --workspace --all-targets -- -D warnings: pass
• cargo +1.98.0 test -p codexbar antigravity -- --test-threads=4: 116 passed, 0 failed (4 new tests: explicit cadence for opaque ids, explicit window over legacy names incl. unknown, blank window fallback, weekly-only Starter)

Affected areas

Rust backend, Antigravity provider only. No shared code, no new dependencies.

UI proof

Not applicable (window minutes only, existing rendering).

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Advanced

Run ID: 37d7f0ed-d25e-4413-a7a8-a5574dbd5788

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[Finesssee]

Thermo-nuclear review

Scope: quota_summary.rs (+82/-4 lines, file is 684 lines, well under 1k). Behavior matches the 0.69.0 spec: an explicit non-empty window replaces the id/name candidates, blank falls back, weekly maps to 10080, 5h to 300, unknown yields no cadence, and the primary/secondary-removed-from-extra_rate_windows invariant is kept. No new dependencies, provider logic stays in the antigravity module, OAuth half not ported (as specified).

Findings (all valid, all small):

1. bucket_kind computed non_empty(bucket.window) twice and built an intermediate names array plus a &[Option<&str>] slice through a match that re-derived the same value. Replace with one match yielding a fixed [Option<&str>; 2].
2. explicit_window_sets_cadence_for_opaque_bucket_ids branched on the expected value inside the loop and used title only as an assert message, so the test logic was harder to read than two direct cases.
3. explicit_window_replaces_legacy_bucket_names used a match minutes inside a loop with a shared trailing assertion; two direct cases with a shared snapshot helper are clearer and drop the duplicated JSON.

Verdict: no correctness or convention problems; fixing the three readability items now, behavior unchanged.

[Finesssee]

Follow-up to the thermo-nuclear review: pushed bf8b80b "Address thermo review".

Fixed: all three findings. bucket_kind now picks its candidates with a single match, and the two explicit-window tests use a shared single_bucket_snapshot helper with direct assertions. Behavior is unchanged.

Left: nothing. Validation on Rust 1.98.0: cargo fmt --all, clippy -p codexbar --all-targets -D warnings clean, cargo test -p codexbar antigravity 116 passed, 0 failed. No UI change.

[Finesssee]

Adversarial validation (lane-A) at bf8b80b

Review verdict: the branch at its published head already matches the 0.69.0 audit row for Antigravity window cadence; local HEAD equals the remote head (bf8b80bc), so no re-merge or fix commit was needed. The thermo-review commit on top of the original port is included in the published head and the checks below were run at exactly that sha.

Checks at bf8b80b (CARGO_TARGET_DIR=W:\cargo-target\lane-a, jobs=4, RUST_TEST_THREADS=4):

• cargo fmt --all --check: pass.
• cargo clippy both manifests --all-targets -- -D warnings: only the 3 documented pre-existing main-drift findings (alibabatokenplan/cli.rs:163, kiro/usage_limits.rs:315, openai/subscription.rs:206); 0 findings in this PR's diff.
• cargo test rust manifest: pass (0 failed / 1 ignored baseline).
• cargo test desktop manifest: only the documented Isolate bootstrap payload test from real settings #684 bootstrap_payload_exposes_every_provider_variant failure (env-dependent baseline on branches without Make the bootstrap catalog test hermetic (#684) #711; expected, hermetic fix lives on release/v0.70.0).

No new code was required; validation-only item.

[Finesssee]

Shipped in v0.70.0: this PR's head is included in main via #735 (merge commit 9d0a37a). Closing as integrated.