feat(spend): add tokscale-compatible local readers for Cursor and Antigravity - #3113
Merged
steipete merged 12 commits intoAug 23, 2026
Merged
Conversation
|
🦞👀 Pull request received. I will update this pull request when review starts. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 9da2dec42f
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Comment on lines
+439
to
+443
| // Provider-specific by design: Cursor local CSV is an offline fallback for the remote dashboard API. | ||
| if provider == .cursor, let local = await self.loadCursorLocalSnapshot( | ||
| now: now, historyDays: clampedHistoryDays) | ||
| { | ||
| return local |
There was a problem hiding this comment.
On macOS, an offline request, expired session, or missing Cursor cookie makes the preceding try await loadRemoteTokenSnapshot throw rather than return nil, so execution never reaches this advertised local fallback even when valid tokscale CSV files exist. Catch the remote Cursor failure, try the local snapshot, and rethrow only if no usable local data is available.
Useful? React with 👍 / 👎.
Comment on lines
+1150
to
+1151
| let filtered = full.data.filter { $0.date >= sinceKey && $0.date <= nowKey } | ||
| let daily = CostUsageDailyReport(data: filtered, summary: full.summary) |
There was a problem hiding this comment.
When the local cache contains rows older than the requested window, filtered correctly limits daily.data, but retaining full.summary makes tokenSnapshot prefer the all-time token and cost totals over those filtered entries. Consequently a 30-day snapshot can report months of spend while exposing only 30 days of rows; recompute the summary from filtered here and in the mirrored Antigravity loader.
Useful? React with 👍 / 👎.
Comment on lines
+84
to
+86
| costUSD: 0, modelsUsed: nil, | ||
| modelBreakdowns: [CostUsageDailyReport.ModelBreakdown( | ||
| modelName: modelId, costUSD: 0, totalTokens: total, requestCount: 1)]) |
There was a problem hiding this comment.
For every nonempty Antigravity cache, each usage entry and model breakdown receives an explicit costUSD: 0, the summary therefore totals zero, and the fetcher labels the result as a list-price estimate. Users with real token usage will consequently see $0 spend rather than an unknown value or an actual model-price estimate; apply the pricing catalog before publishing the snapshot, or leave costs nil until pricing is available.
Useful? React with 👍 / 👎.
Comment on lines
+52
to
+54
| supportsTokenCost: true, | ||
| noDataMessage: { "Antigravity cost summary is not supported." }, | ||
| supportsTokenSnapshot: true), |
There was a problem hiding this comment.
Enabling supportsTokenSnapshot here makes ProviderArchitectureGatekeeperTests.small provider capabilities preserve legacy registries fail on both macOS and Linux because its expected provider sets still omit .antigravity. Update that registry assertion alongside this capability change so the test suite can pass and the new behavior is covered.
AGENTS.md reference: AGENTS.md:L5-L5
Useful? React with 👍 / 👎.
Comment on lines
+177
to
+179
| var c = calendar.dateComponents([.year, .month, .day], from: d) | ||
| c.hour = 12; c.minute = 0; c.second = 0; c.timeZone = TimeZone(secondsFromGMT: 0) | ||
| return calendar.date(from: c) ?? d |
There was a problem hiding this comment.
For a date-only value such as 2026-08-21, the formatter first creates midnight UTC, but extracting components with a negative-offset calendar produces August 20 before the code moves it to noon UTC. Cursor rows are therefore assigned to the previous day for users west of UTC; extract the year/month/day using a UTC calendar or parse the components directly from the date string.
Useful? React with 👍 / 👎.
clawsweeper
Bot
added
merge-risk: 🚨 other
|
Codex review: needs changes before merge. Reviewed August 23, 2026, 9:18 AM ET / 13:18 UTC. ClawSweeper reviewWhat this changesAdds local tokscale CSV and JSONL readers for Cursor and Antigravity and feeds their usage into CodexBar’s spend dashboard. Merge readinessKeep open: the current head still assigns date-only Cursor cache rows using the process calendar rather than CodexBar’s configured reporting calendar, so spend can appear on the wrong day. Priority: P2 Review scores
Verification
Live VerificationCommand: Result: FAIL (failed) — execution before step 1 Assertions:
How this fits togetherCodexBar’s cost-usage fetcher obtains provider usage from remote APIs and local files, normalizes it into daily reports, and supplies those reports to the spend dashboard. This PR adds tokscale cache files as local sources for Cursor and Antigravity. flowchart LR
A[Cursor CSV cache] --> C[Local usage readers]
B[Antigravity JSONL cache] --> C
D[Remote provider fetch] --> E[Cost usage fetcher]
C --> E
E --> F[Daily usage report]
F --> G[Spend dashboard]
Before merge
Findings
Agent review detailsSecurityNone. Review metrics
Merge-risk optionsMaintainer options:
Copy recommended automerge instructionTechnical reviewBest possible solution: Pass the resolved reporting calendar into Cursor CSV parsing and add a cross-time-zone date-only regression test before merging. Do we have a high-confidence way to reproduce the issue? Yes—source reproducible: invoke the package-level token-snapshot path with a configured calendar different from Is this the best way to solve the issue? No—the current implementation is not the best final form because it preserves configured-calendar grouping only after parsing; parsing must receive that same calendar. Full review comments:
Overall correctness: patch is incorrect AGENTS.md: found and applied where relevant. Codex review notes: model internal, reasoning high; reviewed against 4b14ed9c57d3. LabelsLabel justifications:
EvidenceAcceptance criteria:
What I checked:
Likely related people:
Rank-up movesOptional improvements that raise the rating; they are not merge blockers.
Rating scale
Overall follows the weaker of proof and patch quality. Workflow
HistoryReview history (11 earlier review cycles; latest 8 shown)
|
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 79bdb87e2f
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
|
||
| private static func loadAntigravityLocalSnapshot(now: Date, historyDays: Int) async -> CostUsageTokenSnapshot? { | ||
| let report = AntigravityLocalReader.makeDailyReport() | ||
| guard !report.data.isEmpty else { return nil } |
There was a problem hiding this comment.
When no tokscale cache exists or every row is unreadable, returning nil here falls through to CostUsageScanner, whose default branch returns an empty report; loadLocalTokenScanResult then marks every non-Codex report as having established coverage, so tokenSnapshot publishes zero tokens and $0 spend. This makes the optional external cache appear to confirm no usage for Antigravity users who never installed tokscale; stop the Antigravity path or preserve unestablished coverage when the cache is absent.
Useful? React with 👍 / 👎.
| sessionModel = json["modelId"] as? String ?? json["model_id"] as? String | ||
| continue | ||
| } | ||
| if type == "usage" || json["input"] != nil { |
There was a problem hiding this comment.
When the synced cache contains the same response in multiple session JSONLs, this loop counts every copy because it ignores the format's responseId identity and only merges by day. The duplicated response inflates token totals, request counts, and model breakdowns on every refresh; track seen nonempty response IDs across all cache files before aggregating usage records.
Useful? React with 👍 / 👎.
Yuxin-Qiao
force-pushed
the
feat/spend-tokscale-antigravity-cursor
branch
from
August 21, 2026 13:29
79bdb87 to
9c2e542
Compare
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 9c2e542f12
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| @@ -1,245 +0,0 @@ | |||
| // swift-tools-version: 6.2 | |||
There was a problem hiding this comment.
In every checkout of this commit, deleting Package.swift along with all Sources, Tests, workflows, documentation, and nearly every script leaves only Scripts/lint.sh; the package can no longer build or run, and the advertised Cursor/Antigravity implementation is absent. Restore the project tree and apply only the intended reader changes before merging.
AGENTS.md reference: AGENTS.md:L3-L7
Useful? React with 👍 / 👎.
clawsweeper
Bot
added
merge-risk: 🚨 availability
…igravity
- Cursor: read ~/.config/tokscale/cursor-cache/usage*.csv (v1/v2/v3) with
tokstyle column handling, cacheWrite = with-without, noon UTC for date-only,
and CostUsageDailyReport aggregation.
(Sources/CodexBarCore/Providers/Cursor/CursorLocalCSVReader.swift:1)
- Antigravity: read ~/.config/tokscale/antigravity-cache/sessions/*.jsonl
(tokscale JSONL) and stub for ~/.gemini/antigravity-cli/*.db direct SQLite
(ProtoReader to follow). Handles session_meta fallback and dedup.
(Sources/CodexBarCore/Providers/Antigravity/AntigravityLocalReader.swift:1)
- CostUsageFetcher: local fallback before remote for Cursor (offline) and
primary for Antigravity (quota-only before), with Provider-specific by
design comments for gatekeeper.
(Sources/CodexBarCore/CostUsageFetcher.swift:440)
- Antigravity descriptor: enable supportsTokenSnapshot for spend dashboard.
(Sources/CodexBarCore/Providers/Antigravity/AntigravityProviderDescriptor.swift:51)
Reproduced from /tmp/opencodex/src/adapters/cursor/protobuf-events.ts:218
and /tmp/tokscale/crates/tokscale-core/src/sessions/{cursor,antigravity_cli}.rs
Phase 1 of opencodex/tokscale plan, offline-first, no auth.
Yuxin-Qiao
force-pushed
the
feat/spend-tokscale-antigravity-cursor
branch
from
August 21, 2026 19:19
9c2e542 to
b9a4f38
Compare
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: b9a4f38c64
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| self.hasSharedOAuthCredentials(context: context) | ||
| { | ||
| return [app, cli, ide, oauth] | ||
| return [app, cli, ide, oauth, offline] |
There was a problem hiding this comment.
When auto mode has selected or shared OAuth credentials, this ordering places offline after AntigravityOAuthFetchStrategy, but that strategy's shouldFallback always returns false; ProviderFetchPipeline.fetch therefore returns immediately on an expired credential or network error and never attempts the newly added offline strategy. Either place the offline strategy before the terminal OAuth strategy or explicitly permit the intended OAuth-to-offline transition.
Useful? React with 👍 / 👎.
Comment on lines
+43
to
+44
| let cache = self.tokscaleCacheDirectory(home: home) | ||
| return self.countJSONLFiles(in: cache, fileManager: fileManager) |
There was a problem hiding this comment.
When tokscale is configured through TOKSCALE_CONFIG_DIR and there are no Antigravity CLI database files under the default home, countConversations ignores the supplied environment and looks only in ~/.config/tokscale. The local usage reader honors this override, but the offline fetch strategy reports itself unavailable, so users with a valid cache at the configured location do not receive the intended fallback.
Useful? React with 👍 / 👎.
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 773305da2d
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| /// almost immediately after a successor ships. Keep old picker selections routable and | ||
| /// prevent stale payloads from republishing dead wire ids as separate rows. | ||
| private static let retiredFlashTiers: [String: String] = [ | ||
| "gemini-3.6-flash": "gemini-3.7-flash", |
There was a problem hiding this comment.
When Antigravity returns any of these legitimate historical IDs, this table relabels the quota as gemini-3.7-flash and can merge tier-specific entries under that invented identity, so users see a different model than the service reported and potentially lose distinct quota rows. Remove the speculative aliases unless the replacement is an officially released, verified wire ID; repository guidance explicitly prohibits unreleased model names in code and tests.
AGENTS.md reference: AGENTS.md:L22-L22
Useful? React with 👍 / 👎.
| updatedAt: Date(), | ||
| identity: ProviderIdentitySnapshot( | ||
| providerID: .antigravity, | ||
| accountEmail: AntigravitySelectedAccountGuard.selectedAccountEmail(context: context), |
There was a problem hiding this comment.
In explicit CLI mode with a saved OAuth account selected, if the live probes fail and the local cache belongs to a different Antigravity sign-in, this fallback attaches the selected OAuth email even though the conversation files are ambient and contain no account verification. The resulting menu snapshot misattributes the offline count to that saved account; leave the email unset or derive and validate it from local metadata.
Useful? React with 👍 / 👎.
Yuxin-Qiao
force-pushed
the
feat/spend-tokscale-antigravity-cursor
branch
from
August 21, 2026 20:03
773305d to
ba5a4db
Compare
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: ba5a4db494
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| costUSD: 0, | ||
| modelsUsed: nil, | ||
| modelBreakdowns: []) | ||
| let tot = row.input + row.output + row.cacheRead + row.cacheWrite |
There was a problem hiding this comment.
When cache buckets are present, this recomputes a total that can disagree with the exported Total Tokens value and inflates every daily/model/summary total. For example, the newly added v1 fixture declares 1,500 total tokens, but this expression produces 1,900 (800 + 300 + 400 + 400); parse and aggregate the format's total-token column instead of deriving a different value.
Useful? React with 👍 / 👎.
| noDataMessage: { "Antigravity cost summary is not supported." }), | ||
| supportsTokenCost: true, | ||
| noDataMessage: { "Antigravity cost summary is not supported." }, | ||
| supportsTokenSnapshot: true), |
There was a problem hiding this comment.
On Linux this unconditional capability makes small provider capabilities preserve legacy registries fail because its #else expectation still omits .antigravity. Fresh evidence after the earlier review comment is that this commit updated only the macOS expectation while leaving the Linux set unchanged, even though the descriptor is not platform-gated.
Useful? React with 👍 / 👎.
clawsweeper
Bot
added
merge-risk: 🚨 other
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 667fe62545
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Comment on lines
+1145
to
+1146
| let full = CursorLocalCSVReader.makeDailyReport(from: allRows, now: now) | ||
| let cal = Calendar.current |
There was a problem hiding this comment.
When the pinned cost-usage timezone differs from Calendar.current—for example, after the user travels—both new local fallbacks ignore the calendar supplied through CostUsageFetcher and instead parse, group, filter, and resolve the current session using the process timezone. Midnight-adjacent Cursor or Antigravity events can therefore move to another day or fall across the requested-window boundary, violating the pinned-bucket behavior; pass the configured scanner calendar through both readers, filtering, and tokenSnapshot.
Useful? React with 👍 / 👎.
clawsweeper
Bot
removed
rating: 🧂 unranked krab
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: aa464e6b6d
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| let daily = CostUsageDailyReport(data: filtered, summary: filteredSummary) | ||
| return Self.tokenSnapshot( | ||
| from: daily, | ||
| now: now, |
There was a problem hiding this comment.
When the tokscale cache has not synced recently, passing refresh-time now here and leaving historyCoverageIsEstablished at its default makes stale local data appear freshly updated with complete coverage. If every cached Cursor row predates the requested window, this returns an established empty snapshot and reports zero current spend instead of indicating stale/unavailable data; the mirrored Antigravity return has the same problem. Derive updatedAt and coverage from the cache files or decline the fallback when it cannot establish the requested window.
Useful? React with 👍 / 👎.
Comment on lines
+187
to
+189
| var c = utcCal.dateComponents([.year, .month, .day], from: d) | ||
| c.hour = 12; c.minute = 0; c.second = 0; c.timeZone = TimeZone(secondsFromGMT: 0) | ||
| return utcCal.date(from: c) ?? d |
There was a problem hiding this comment.
For users in UTC+13 or UTC+14, representing a date-only CSV value as noon UTC moves it into the following local day when makeDailyReport derives the configured-calendar key. For example, 2026-08-21 becomes August 22 in Pacific/Kiritimati, shifting totals and potentially crossing the requested-window boundary. Preserve the parsed day components directly in the target calendar rather than routing a date-only value through a UTC instant.
Useful? React with 👍 / 👎.
| let firstStamp = Int64(noon.timeIntervalSince1970 * 1000) | ||
| let secondStamp = firstStamp + 3_600_000 | ||
| let jsonl = [ | ||
| #"{"type":"session_meta","modelId":"gemini-3.7-flash"}"#, |
There was a problem hiding this comment.
The new Antigravity fixture hard-codes gemini-3.7-flash, which is neither a released model nor clearly fictitious. Replace it and the matching expectation with a released or unmistakably synthetic identifier so tests do not expose an unreleased model name.
AGENTS.md reference: AGENTS.md:L22-L22
Useful? React with 👍 / 👎.
Contributor
Author
|
@clawsweeper re-review |
|
🦞🧹 I asked ClawSweeper to review this item again. |
- calendar before now in makeDailyReport - implicit optional init - wrap long lines and andOperator
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 57e4d9e83d
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Comment on lines
+1185
to
+1189
| from: daily, | ||
| now: now, | ||
| historyDays: historyDays, | ||
| useCurrentLocalDayForSession: true, | ||
| costProvenance: .listPriceEstimate) |
There was a problem hiding this comment.
When the configured scanner calendar differs from Calendar.current near a day boundary, this call builds daily keys with the configured calendar but lets tokenSnapshot search for today's entry using its default calendar, so current-session tokens and cost can incorrectly become zero. Fresh evidence after the earlier calendar comment is that the loader now receives and uses fallbackCalendar for parsing and filtering, but this final call—and the mirrored Antigravity call—still omits calendar: calendar.
Useful? React with 👍 / 👎.
clawsweeper
Bot
added
rating: 🦪 silver shellfish
Contributor
Author
|
Added terminal-only real behavior proof to the PR body (no screenshots per request). Synthetic Shows: @clawsweeper re-review |
|
🦞🧹 I asked ClawSweeper to review this item again. |
clawsweeper
Bot
added
rating: 🦐 gold shrimp
…-only, fixture model - Preserve cache freshness: return nil when filtered window is empty instead of publishing established zero with now timestamp - Pass pinned calendar into tokenSnapshot for Cursor/Antigravity local snapshots - Keep date-only Cursor CSV rows in configured calendar's noon, not UTC noon - Use clearly fictitious test model test-model-antigravity-a
Contributor
Author
|
Fixed 4 review blockers at 075eac7 and added production fetcher proof to PR body.
Production fetcher @clawsweeper re-review |
|
🦞👀 Re-review progress:
|
Contributor
Author
|
Head 60fdee7: fixed lint (test-model-a, line length) — @clawsweeper re-review |
|
🦞🧹 I asked ClawSweeper to review this item again. |
clawsweeper
Bot
added
proof: sufficient
Allowlist lines 1339->1335 and 1695->1691 after 075eac7 freshness/calendar fixes
steipete
added a commit
that referenced
this pull request
Aug 23, 2026
steipete
added a commit
that referenced
this pull request
Aug 23, 2026
…easoning split, stale (#3120) * Align Codex token parsing with tokscale stale snapshots - skip lightly regressed cumulative snapshots before interleaved latching - take the maximum of cached and cache-read fields in all parsers - cover cache field selection and out-of-order snapshot accounting with focused tests * Refresh Codex parser hash * Parse bare usage rows in Codex rollouts * Fix stale reasoning and fallback cache parity * Fix Codex fallback test fixture line handling * fix(antigravity): repair offline fallback proof and oauth fallback; fix(spend): limit concurrent dashboard fetches to 3 * fix(lint): break long lines in offline fallback proof tests * fix(tests): update gatekeeper anchors for spend dashboard concurrency limit * fix(tests): correct gatekeeper line anchors for concurrent dashboard fix * docs: update appcast for 0.54.1 * chore: open 0.54.2 unreleased changelog section * Stop re-merging the Codex plan-utilization history with itself on every refresh (#3141) `materializeCodexPlanUtilizationHistoryIfNeeded` exists to fold legacy, opaque and unscoped Codex plan-utilization buckets into the canonical account bucket. Its scoped loop also appended the canonical bucket's own histories to `historiesToMerge` — `matchesTargetContinuity` is true for `rawKey == canonicalKey`, and only the removal of the old key was guarded — so `guard !historiesToMerge.isEmpty` never fired once the canonical bucket had any history, and the migration merge ran on every successful provider refresh and every menu open, merging the history with itself. That merge is quadratic: `updatedPlanUtilizationEntries` copied the whole entry array per entry, scanned it linearly for the insertion point, and allocated the same-hour slice. Measured with an optimized standalone reproduction over a real three-month-old history (session 1909 entries, weekly 2239): 20.6 ms of MainActor time per call, scaling ~3.9x per doubling. `planUtilizationMaxSamples` allows 17520 entries per series, so it would keep growing. Two changes: - Track whether a foreign source actually contributed and require that in the guard, so the canonical-only case returns without merging or rewriting anything. Every path where a legacy, opaque or unscoped bucket contributes is untouched; `legacyRawKeysToRemove` is populated only in branches that also set the flag, so no removal is skipped, and `providerBuckets.unscoped` is cleared only inside the branch that sets it. - Make the merge itself near-linear: `updatedPlanUtilizationEntries` mutates the array in place and finds the insertion point with a binary search for the same strict upper bound (with a fast path for the common append), and `mergedPlanUtilizationHistories` accumulates per series and builds each history once. The binary search assumes entries are sorted by `capturedAt`, which every in-app producer guaranteed through `PlanUtilizationSeriesHistory`'s designated initializer — except the synthesized `Codable` decoder, which assigned entries verbatim from JSON. An explicit `init(from:)` now routes decoding through that initializer, so an on-disk history written by an older build or edited by hand cannot smuggle in an unsorted series. The skipped self-merge also incidentally re-canonicalized per-hour peaks on read; that repair belongs at load time, not on every refresh, and is not reintroduced here. The visible effect is that at most one extra real observation per affected hour is kept. Tests: canonical-only history is returned untouched and enqueues no persistence write (the history revision is unchanged); a genuine foreign merge matches an explicit expected result across overlapping hours, out-of-order sources, distinct series and retention trimming; the binary search's upper-bound contract is pinned directly through a DEBUG shim over an array with a run of equal timestamps (a lower bound would return a different index); and decoding a series whose JSON entries are out of order yields a sorted series. Implemented by grok-4.6 (xhigh) via implementation-loop; reviewed hunk by hunk plus an independent deep review that confirmed both equivalences by differential fuzzing (200k sorted cases with no mismatch) and found the decoder gap, fixed in one iterate round. Gatekeeper line anchors for the touched file were re-verified independently. The DEBUG sortedness assertion is checked once per merged series rather than once per inserted entry: a per-entry check is itself O(n) and reintroduced, in debug builds, exactly the quadratic scan this insertion path removes (measured over the real 4160-entry history: a legacy migration took ~1000 ms with the per-entry assertion versus ~15 ms without it). Co-authored-by: Claude Fable 5 <noreply@anthropic.com> * Fix Codex day cost blanked by trace-only priority turns (#3150) Row ownership evidence compared the retained rows against the persisted standard/priority split using the trace database's tier classification. The persisted maps come from the rows' own pricingMode, so a turn the trace reports as priority after its rows were persisted as standard read as a row-ownership mismatch, the rows lost trust, and the day fell back to the aggregate — which returns nil for long-context tiered models, so the whole day's cost disappeared from the menu, the chart and the window total. Judge retention against both classifications and flag only a group that matches neither. A wrongly retained row set still fails both, because the persisted totals are canonical for the file and tier classification never changes how many tokens the rows carry. Co-authored-by: Claude Opus 5 <noreply@anthropic.com> * docs: credit #3141 and #3150 changelog entries * fix(qwen-cloud): restore Brave browser support in cookie import (#3148) * fix(qwen-cloud): restore Brave browser support, narrowed to Chrome+Brave per AGENTS.md Qwen Cloud's cookie import was restricted to [.chrome] only (commit 529cc6c 'Keep Qwen imports Chrome-only'). Brave users hit 'No Qwen Cloud session cookies found in browsers' even when they had a valid Qwen Cloud session in Brave, because their cookies were never probed. This commit restores Brave in the import order, but follows AGENTS.md L48 ('default Chrome-only when possible to avoid other browser prompts; override via browser list when needed'). The override is the minimum necessary: Chrome + Brave. The other Chromium browsers (chromeBeta, edge, arc, firefox, safari) are deliberately omitted to avoid unsolicited Keychain / browser-store access prompts on automatic refreshes from browsers that don't carry a Qwen Cloud session. Brave is kept because it shares the same Chromium Safe Storage format as Chrome and is a common Qwen Cloud authentication target. Also adds docs/qwen-cloud-proof/README.md with the redacted end-to-end proof captured against the live Qwen Cloud API from the user's Mac after granting the modified binary access to 'Brave Safe Storage' in macOS Keychain. * fix(qwen-cloud): recovery message now names Brave alongside Chrome ClawSweeper P2 follow-up on #3148: when the Brave cookie import fails, QwenCloudSettingsError.missingCookie's recovery message still told users to sign in to Chrome and grant access to Chrome Safe Storage. Now that Brave is a supported source, the message must name both browsers and their respective Safe Storage entries, otherwise a Brave-only user would be told to use Chrome and never find the working path. Updates the error description to: 'No Qwen Cloud session cookies found in browsers. Sign in to Qwen Cloud in Chrome or Brave, allow CodexBar to access the corresponding Safe Storage in Keychain Access (Chrome Safe Storage and/or Brave Safe Storage), or paste a manual Cookie header.' Adds focused test coverage: - missing cookie error mentions both supported browsers and their safe storage - missing cookie error appends non-empty details - missing cookie error omits empty details 35/35 Qwen Cloud tests pass (32 prior + 3 new). * Fix OpenRouter completed-day activity query (#3138) * Preserve unknown Grok period usage (#3159) Co-authored-by: anupamchugh <8416306+anupamchugh@users.noreply.github.com> * fix: report non-writable CLI path conflicts (#3153) * fix: prefer successful CLI install status * fix: keep CLI path conflicts visible * fix: report non-writable CLI path conflicts * docs: add CLI conflict behavior proof * docs: add CLI install comparison screenshots * Fix single-quota icon scaling (#3155) * docs: credit #3138 #3148 #3153 #3155 #3159 changelog entries * fix(spend): silent refresh and invalidation coverage (#3106) * fix(spend): bucket calendar for all heatmap dates and full revision hash - SpendActivityDateFormatting.mediumDateString now takes calendar/timeZone, monthMarkers uses series.calendar, tooltips and accessibility use bucket calendar. - selectedDay renormalized on calendar change to keep toggle correct. - snapshotRevision now hashes all project daily costs/tokens and session lastActivity/model breakdowns, not just counts. Fixes ClawSweeper P2 for #3106. * fix(gatekeeper): update anchors and add provider-specific design markers for spend dashboard * Update provider gatekeeper anchors for v0.54 rebase * fix(test): pin claude spend snapshot in observation test * fix(test): seed pinned claude spend publication before first snapshot * fix: resolve remaining conflict markers from gatekeeper rebase * fix(lint): shorten Sakana test lines * fix(spend): restore heatmap calendar property lost in rebase * Extend spend publication test wait * Restore spend gatekeeper anchors after rebase * fix(spend): sync independent snapshot and bucket calendar normalization for 3106 - publishSpendDashboardTokenSnapshotState now calls synchronizeSharedSpendDashboardAfterTokenPublication - heatmap calendar onChange no longer renormalizes selectedDay via stale controller - SpendDashboardController.update now normalizes selectedDay atomically when bucketTimeZoneIdentifier changes - update gatekeeper anchors for shifted lines (1620,1649,1666,1693) * test(spend): cover independent snapshot sync for 3106 Exercise the direct independent publication path added at UsageStore+SpendDashboardTokenCost.swift:181. The prior focused test seeded Claude before observation and then used the regular Codex publisher, which already syncs independently, so removing that line would not fail. Add a post-start Claude snapshot via _setSpendDashboardTokenSnapshotForTesting and assert the shared dashboard debounced sync is scheduled and the publication inputs update. Verified: swiftformat clean, swiftlint --strict clean, swift test --filter SpendDashboardPublicationTests 18 tests passed. * docs: add fresh-bundle proof for 3106 Add redacted menu-icon crop and dashboard snapshot from debug build 2798eec (swift build --target CodexBarCLI, .build/debug/CodexBarCLI dashboard --pretty). The snapshot shows the shared spend controller produces a dashboard with provider rows/windows, confirming the independent-sync and calendar paths are live in the fresh binary. * docs: add menu and Spend dashboard screenshots for 3106 Add redacted screenshots from fresh debug build 45ba984: - 3106-menu-after-fix.png: menu bar extra open, showing provider rows - 3106-settings-after-fix.png: Settings window (general) - 3106-spend-dashboard-after-fix.png: Usage & Spend pane (usageSpend) with heatmap and Overview, confirming the shared controller renders in the fresh bundle. * docs: remove screenshots for 3106 per request Keep only the redacted CLI dashboard snapshot JSON as fresh-bundle proof; screenshots are not needed. * Improve Antigravity retrieval: retired Flash alias and offline fallback (#3119) * fix(antigravity): allow OAuth errors to fallback to offline when local data exists Fix P2 from Codex review on #3119: AntigravityOAuthFetchStrategy.shouldFallback now checks hasOfflineData, so expired credentials do not block offline. * fix(antigravity): unbind offline account, read app-data, bound scans + proof - Offline snapshot now has nil accountEmail (P1) - OfflineStore also counts $HOME/.gemini/antigravity and .../conversations (P2) - SpendDashboardController bounds Codex scans to 3 concurrent (P2) - Add AntigravityOfflineFallbackProofTests covering app-data and nil email * fix: remove broken proof test, keep P1/P2 fixes and shell proof * fix(gatekeeper): update SpendDashboardController anchors after bounding Codex scans * fix: revert bounded Codex scans (keep offline P1/P2), restore gatekeeper * feat(spend): add tokscale-compatible local readers for Cursor and Antigravity (#3113) * feat(spend): add tokscale-compatible local readers for Cursor and Antigravity - Cursor: read ~/.config/tokscale/cursor-cache/usage*.csv (v1/v2/v3) with tokstyle column handling, cacheWrite = with-without, noon UTC for date-only, and CostUsageDailyReport aggregation. (Sources/CodexBarCore/Providers/Cursor/CursorLocalCSVReader.swift:1) - Antigravity: read ~/.config/tokscale/antigravity-cache/sessions/*.jsonl (tokscale JSONL) and stub for ~/.gemini/antigravity-cli/*.db direct SQLite (ProtoReader to follow). Handles session_meta fallback and dedup. (Sources/CodexBarCore/Providers/Antigravity/AntigravityLocalReader.swift:1) - CostUsageFetcher: local fallback before remote for Cursor (offline) and primary for Antigravity (quota-only before), with Provider-specific by design comments for gatekeeper. (Sources/CodexBarCore/CostUsageFetcher.swift:440) - Antigravity descriptor: enable supportsTokenSnapshot for spend dashboard. (Sources/CodexBarCore/Providers/Antigravity/AntigravityProviderDescriptor.swift:51) Reproduced from /tmp/opencodex/src/adapters/cursor/protobuf-events.ts:218 and /tmp/tokscale/crates/tokscale-core/src/sessions/{cursor,antigravity_cli}.rs Phase 1 of opencodex/tokscale plan, offline-first, no auth. * test(readers): cover cursor csv schemas and antigravity cache fallback * fix(test): include antigravity in cost capable dashboard sources * fix(spend): honor CSV total tokens and add Antigravity Linux capability * fix(test): honor cursor CSV total tokens column in aggregation * fix(spend): repair 3113 tokscale readers P1s - catch remote Cursor errors before falling back to local CSV - recompute summaries after window filtering for Cursor and Antigravity - keep Antigravity costs nil (unpriced) and deduplicate by responseId - parse date-only CSV rows with UTC calendar - thread fallback calendar through loaders * fix(lint): repair 3113 build and format - calendar before now in makeDailyReport - implicit optional init - wrap long lines and andOperator * style: swiftformat wrap for 3113 * Fix 3113 provider gatekeeper anchors * fix(spend): address 3113 review findings -- freshness, calendar, date-only, fixture model - Preserve cache freshness: return nil when filtered window is empty instead of publishing established zero with now timestamp - Pass pinned calendar into tokenSnapshot for Cursor/Antigravity local snapshots - Keep date-only Cursor CSV rows in configured calendar's noon, not UTC noon - Use clearly fictitious test model test-model-antigravity-a * style: fix line length for fixture model * fix(test): update gatekeeper anchors for CostUsageFetcher line drift Allowlist lines 1339->1335 and 1695->1691 after 075eac7 freshness/calendar fixes * test: repair gatekeeper anchors and regenerate parser hash on merged tree --------- Co-authored-by: Yuxin-Qiao <2242016570@qq.com> Co-authored-by: Peter Steinberger <steipete@gmail.com> Co-authored-by: Olddonkey <olddonkeyblog@gmail.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> Co-authored-by: Umut Keltek <35880258+umutkeltek@users.noreply.github.com> Co-authored-by: kiranmagic7 <kiranmagic@proton.me> Co-authored-by: Anupam Chugh <anupam.chugh@gmail.com> Co-authored-by: anupamchugh <8416306+anupamchugh@users.noreply.github.com> Co-authored-by: yicone <yicone@gmail.com> Co-authored-by: Akshay Prabhu <12824090+akshayprabhu200@users.noreply.github.com>
steipete
added a commit
that referenced
this pull request
Aug 24, 2026
…ay reloads (#3136) * Price OpenCodex usage once per entry and stop per-entry catalog/overlay reloads The OpenCodex spend source (`~/.opencodex/usage.jsonl` → `OpenCodexUsageFanOut` → `OpenCodexUsageAggregator.snapshot`) re-resolved pricing context per entry: `listPriceUSD` called `CostUsagePricing.codexCostUSD` without a pre-resolved models.dev catalog, so every call went through `ModelsDevCache.load` → `FileManager.attributesOfItem` (a stat plus an extended-attribute read), and without a pre-resolved custom-pricing overlay, so every call also re-read the overlay file location. Each windowed entry was priced three times (day, session and hour accumulators), and day keys / hour buckets were recomputed through Calendar per entry. On a 35k-entry log (all inside the 30-day window) that is ~100k stat+xattr syscalls and ~70k Calendar interval computations per refresh — in the running app this was the 25–35 s CPU spike on every adaptive refresh (sampled: `snapshotsBySubscription` → `attributesOfItem` → `getxattr`/`listxattr`). Changes (snapshot output is byte-identical; verified against a reference implementation in tests and by diffing CLI JSON on frozen inputs): - Resolve the models.dev catalog and the custom-pricing overlay once per fan-out / snapshot and pass them down; price each windowed entry once and reuse the value for the day/session/hour/model merges. A missing catalog is substituted with an empty catalog so the degraded path never falls back to per-call loads. - Memoize the local-day key and hour-bucket start per calendar interval using the calendar's own `[start, end)` intervals (DST-correct; no 86400/3600 arithmetic). - `ModelsDevCache.load` reads (mtime, size) via POSIX `stat` instead of `attributesOfItem` (which also reads xattrs); memo/invalidation semantics unchanged. This helps every caller repo-wide. CodexParserHash is regenerated because ModelsDevPricing.swift is in the hashed set; the previous hash (3c984b655688593f) is added to compatiblePredecessorParserHashes since parsing and the persisted row shape are unchanged, so existing cost-usage.sqlite stores are adopted on upgrade instead of rebuilt. Measured (release CodexBarCLI, isolated cache root, real 41.7 MB / ~35k-entry usage.jsonl, same machine, `cost --provider codex --days 30`), OpenCodex path isolated with identical frozen inputs: - OpenCodex path alone (empty codex home, identical frozen inputs, CLI JSON output identical apart from `updatedAt`): cold 14.3 s real / 9.1 s user / 4.9 s sys / 193 G instructions → 2.6 s / 2.4 s / 0.1 s / 40 G warm (store cache hit) 13.8 s / 8.3 s / 5.3 s / 166 G → 1.2 s / 1.1 s / 0.04 s / 13 G - Full `cost --provider codex` CLI run on live data, steady state after the log grew (the app's per-refresh case): ~11 s → ~3.5 s real (7.3–9.2 s → 3.2 s user); cold 26 s → 14 s. Peak footprint unchanged (~430 MB cold/grown, ~120–140 MB warm). Peak memory is unchanged — the remaining transient is the append-only log re-parse (`OpenCodexUsageStore` identity = path|size|mtime), left for a follow-up. Tests: equivalence against an independent reference implementation (mixed providers, estimated/reported/unreported/unsupported, custom overlay, duplicate request IDs, DST transitions in America/Los_Angeles and America/Santiago), metadata-read counting proving one catalog load per snapshot (zero with an injected catalog), day/hour memo boundary cases, and ModelsDevCache memo invalidation on size/mtime change after the stat switch. Implemented by grok-4.6 (xhigh) via implementation-loop; reviewed hunk by hunk plus an independent deep review; one iterate round. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * docs: add changelog entry for #3136 Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * docs: comment the OpenCodex price-once context and memo semantics Explain why the models.dev catalog and the custom-pricing overlay are resolved once per snapshot / fan-out, why a missing catalog is substituted with an empty one (so the degraded path never falls back to per-call ModelsDevCache.load), the two-level overlay precedence in listPriceUSD, why the day-key memo cannot disagree with CostUsageLocalDay.key, and that the metadata-read recorder is task-local test-only instrumentation. Comments only; CodexParserHash is regenerated because ModelsDevPricing.swift is in the hashed set (no shipped hash is affected; the predecessor list is unchanged). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * docs: update appcast for 0.54.1 * chore: open 0.54.2 unreleased changelog section * Stop re-merging the Codex plan-utilization history with itself on every refresh (#3141) `materializeCodexPlanUtilizationHistoryIfNeeded` exists to fold legacy, opaque and unscoped Codex plan-utilization buckets into the canonical account bucket. Its scoped loop also appended the canonical bucket's own histories to `historiesToMerge` — `matchesTargetContinuity` is true for `rawKey == canonicalKey`, and only the removal of the old key was guarded — so `guard !historiesToMerge.isEmpty` never fired once the canonical bucket had any history, and the migration merge ran on every successful provider refresh and every menu open, merging the history with itself. That merge is quadratic: `updatedPlanUtilizationEntries` copied the whole entry array per entry, scanned it linearly for the insertion point, and allocated the same-hour slice. Measured with an optimized standalone reproduction over a real three-month-old history (session 1909 entries, weekly 2239): 20.6 ms of MainActor time per call, scaling ~3.9x per doubling. `planUtilizationMaxSamples` allows 17520 entries per series, so it would keep growing. Two changes: - Track whether a foreign source actually contributed and require that in the guard, so the canonical-only case returns without merging or rewriting anything. Every path where a legacy, opaque or unscoped bucket contributes is untouched; `legacyRawKeysToRemove` is populated only in branches that also set the flag, so no removal is skipped, and `providerBuckets.unscoped` is cleared only inside the branch that sets it. - Make the merge itself near-linear: `updatedPlanUtilizationEntries` mutates the array in place and finds the insertion point with a binary search for the same strict upper bound (with a fast path for the common append), and `mergedPlanUtilizationHistories` accumulates per series and builds each history once. The binary search assumes entries are sorted by `capturedAt`, which every in-app producer guaranteed through `PlanUtilizationSeriesHistory`'s designated initializer — except the synthesized `Codable` decoder, which assigned entries verbatim from JSON. An explicit `init(from:)` now routes decoding through that initializer, so an on-disk history written by an older build or edited by hand cannot smuggle in an unsorted series. The skipped self-merge also incidentally re-canonicalized per-hour peaks on read; that repair belongs at load time, not on every refresh, and is not reintroduced here. The visible effect is that at most one extra real observation per affected hour is kept. Tests: canonical-only history is returned untouched and enqueues no persistence write (the history revision is unchanged); a genuine foreign merge matches an explicit expected result across overlapping hours, out-of-order sources, distinct series and retention trimming; the binary search's upper-bound contract is pinned directly through a DEBUG shim over an array with a run of equal timestamps (a lower bound would return a different index); and decoding a series whose JSON entries are out of order yields a sorted series. Implemented by grok-4.6 (xhigh) via implementation-loop; reviewed hunk by hunk plus an independent deep review that confirmed both equivalences by differential fuzzing (200k sorted cases with no mismatch) and found the decoder gap, fixed in one iterate round. Gatekeeper line anchors for the touched file were re-verified independently. The DEBUG sortedness assertion is checked once per merged series rather than once per inserted entry: a per-entry check is itself O(n) and reintroduced, in debug builds, exactly the quadratic scan this insertion path removes (measured over the real 4160-entry history: a legacy migration took ~1000 ms with the per-entry assertion versus ~15 ms without it). Co-authored-by: Claude Fable 5 <noreply@anthropic.com> * Fix Codex day cost blanked by trace-only priority turns (#3150) Row ownership evidence compared the retained rows against the persisted standard/priority split using the trace database's tier classification. The persisted maps come from the rows' own pricingMode, so a turn the trace reports as priority after its rows were persisted as standard read as a row-ownership mismatch, the rows lost trust, and the day fell back to the aggregate — which returns nil for long-context tiered models, so the whole day's cost disappeared from the menu, the chart and the window total. Judge retention against both classifications and flag only a group that matches neither. A wrongly retained row set still fails both, because the persisted totals are canonical for the file and tier classification never changes how many tokens the rows carry. Co-authored-by: Claude Opus 5 <noreply@anthropic.com> * docs: credit #3141 and #3150 changelog entries * fix(qwen-cloud): restore Brave browser support in cookie import (#3148) * fix(qwen-cloud): restore Brave browser support, narrowed to Chrome+Brave per AGENTS.md Qwen Cloud's cookie import was restricted to [.chrome] only (commit 529cc6c 'Keep Qwen imports Chrome-only'). Brave users hit 'No Qwen Cloud session cookies found in browsers' even when they had a valid Qwen Cloud session in Brave, because their cookies were never probed. This commit restores Brave in the import order, but follows AGENTS.md L48 ('default Chrome-only when possible to avoid other browser prompts; override via browser list when needed'). The override is the minimum necessary: Chrome + Brave. The other Chromium browsers (chromeBeta, edge, arc, firefox, safari) are deliberately omitted to avoid unsolicited Keychain / browser-store access prompts on automatic refreshes from browsers that don't carry a Qwen Cloud session. Brave is kept because it shares the same Chromium Safe Storage format as Chrome and is a common Qwen Cloud authentication target. Also adds docs/qwen-cloud-proof/README.md with the redacted end-to-end proof captured against the live Qwen Cloud API from the user's Mac after granting the modified binary access to 'Brave Safe Storage' in macOS Keychain. * fix(qwen-cloud): recovery message now names Brave alongside Chrome ClawSweeper P2 follow-up on #3148: when the Brave cookie import fails, QwenCloudSettingsError.missingCookie's recovery message still told users to sign in to Chrome and grant access to Chrome Safe Storage. Now that Brave is a supported source, the message must name both browsers and their respective Safe Storage entries, otherwise a Brave-only user would be told to use Chrome and never find the working path. Updates the error description to: 'No Qwen Cloud session cookies found in browsers. Sign in to Qwen Cloud in Chrome or Brave, allow CodexBar to access the corresponding Safe Storage in Keychain Access (Chrome Safe Storage and/or Brave Safe Storage), or paste a manual Cookie header.' Adds focused test coverage: - missing cookie error mentions both supported browsers and their safe storage - missing cookie error appends non-empty details - missing cookie error omits empty details 35/35 Qwen Cloud tests pass (32 prior + 3 new). * Fix OpenRouter completed-day activity query (#3138) * Preserve unknown Grok period usage (#3159) Co-authored-by: anupamchugh <8416306+anupamchugh@users.noreply.github.com> * fix: report non-writable CLI path conflicts (#3153) * fix: prefer successful CLI install status * fix: keep CLI path conflicts visible * fix: report non-writable CLI path conflicts * docs: add CLI conflict behavior proof * docs: add CLI install comparison screenshots * Fix single-quota icon scaling (#3155) * docs: credit #3138 #3148 #3153 #3155 #3159 changelog entries * fix(spend): silent refresh and invalidation coverage (#3106) * fix(spend): bucket calendar for all heatmap dates and full revision hash - SpendActivityDateFormatting.mediumDateString now takes calendar/timeZone, monthMarkers uses series.calendar, tooltips and accessibility use bucket calendar. - selectedDay renormalized on calendar change to keep toggle correct. - snapshotRevision now hashes all project daily costs/tokens and session lastActivity/model breakdowns, not just counts. Fixes ClawSweeper P2 for #3106. * fix(gatekeeper): update anchors and add provider-specific design markers for spend dashboard * Update provider gatekeeper anchors for v0.54 rebase * fix(test): pin claude spend snapshot in observation test * fix(test): seed pinned claude spend publication before first snapshot * fix: resolve remaining conflict markers from gatekeeper rebase * fix(lint): shorten Sakana test lines * fix(spend): restore heatmap calendar property lost in rebase * Extend spend publication test wait * Restore spend gatekeeper anchors after rebase * fix(spend): sync independent snapshot and bucket calendar normalization for 3106 - publishSpendDashboardTokenSnapshotState now calls synchronizeSharedSpendDashboardAfterTokenPublication - heatmap calendar onChange no longer renormalizes selectedDay via stale controller - SpendDashboardController.update now normalizes selectedDay atomically when bucketTimeZoneIdentifier changes - update gatekeeper anchors for shifted lines (1620,1649,1666,1693) * test(spend): cover independent snapshot sync for 3106 Exercise the direct independent publication path added at UsageStore+SpendDashboardTokenCost.swift:181. The prior focused test seeded Claude before observation and then used the regular Codex publisher, which already syncs independently, so removing that line would not fail. Add a post-start Claude snapshot via _setSpendDashboardTokenSnapshotForTesting and assert the shared dashboard debounced sync is scheduled and the publication inputs update. Verified: swiftformat clean, swiftlint --strict clean, swift test --filter SpendDashboardPublicationTests 18 tests passed. * docs: add fresh-bundle proof for 3106 Add redacted menu-icon crop and dashboard snapshot from debug build 2798eec (swift build --target CodexBarCLI, .build/debug/CodexBarCLI dashboard --pretty). The snapshot shows the shared spend controller produces a dashboard with provider rows/windows, confirming the independent-sync and calendar paths are live in the fresh binary. * docs: add menu and Spend dashboard screenshots for 3106 Add redacted screenshots from fresh debug build 45ba984: - 3106-menu-after-fix.png: menu bar extra open, showing provider rows - 3106-settings-after-fix.png: Settings window (general) - 3106-spend-dashboard-after-fix.png: Usage & Spend pane (usageSpend) with heatmap and Overview, confirming the shared controller renders in the fresh bundle. * docs: remove screenshots for 3106 per request Keep only the redacted CLI dashboard snapshot JSON as fresh-bundle proof; screenshots are not needed. * Improve Antigravity retrieval: retired Flash alias and offline fallback (#3119) * fix(antigravity): allow OAuth errors to fallback to offline when local data exists Fix P2 from Codex review on #3119: AntigravityOAuthFetchStrategy.shouldFallback now checks hasOfflineData, so expired credentials do not block offline. * fix(antigravity): unbind offline account, read app-data, bound scans + proof - Offline snapshot now has nil accountEmail (P1) - OfflineStore also counts $HOME/.gemini/antigravity and .../conversations (P2) - SpendDashboardController bounds Codex scans to 3 concurrent (P2) - Add AntigravityOfflineFallbackProofTests covering app-data and nil email * fix: remove broken proof test, keep P1/P2 fixes and shell proof * fix(gatekeeper): update SpendDashboardController anchors after bounding Codex scans * fix: revert bounded Codex scans (keep offline P1/P2), restore gatekeeper * feat(spend): add tokscale-compatible local readers for Cursor and Antigravity (#3113) * feat(spend): add tokscale-compatible local readers for Cursor and Antigravity - Cursor: read ~/.config/tokscale/cursor-cache/usage*.csv (v1/v2/v3) with tokstyle column handling, cacheWrite = with-without, noon UTC for date-only, and CostUsageDailyReport aggregation. (Sources/CodexBarCore/Providers/Cursor/CursorLocalCSVReader.swift:1) - Antigravity: read ~/.config/tokscale/antigravity-cache/sessions/*.jsonl (tokscale JSONL) and stub for ~/.gemini/antigravity-cli/*.db direct SQLite (ProtoReader to follow). Handles session_meta fallback and dedup. (Sources/CodexBarCore/Providers/Antigravity/AntigravityLocalReader.swift:1) - CostUsageFetcher: local fallback before remote for Cursor (offline) and primary for Antigravity (quota-only before), with Provider-specific by design comments for gatekeeper. (Sources/CodexBarCore/CostUsageFetcher.swift:440) - Antigravity descriptor: enable supportsTokenSnapshot for spend dashboard. (Sources/CodexBarCore/Providers/Antigravity/AntigravityProviderDescriptor.swift:51) Reproduced from /tmp/opencodex/src/adapters/cursor/protobuf-events.ts:218 and /tmp/tokscale/crates/tokscale-core/src/sessions/{cursor,antigravity_cli}.rs Phase 1 of opencodex/tokscale plan, offline-first, no auth. * test(readers): cover cursor csv schemas and antigravity cache fallback * fix(test): include antigravity in cost capable dashboard sources * fix(spend): honor CSV total tokens and add Antigravity Linux capability * fix(test): honor cursor CSV total tokens column in aggregation * fix(spend): repair 3113 tokscale readers P1s - catch remote Cursor errors before falling back to local CSV - recompute summaries after window filtering for Cursor and Antigravity - keep Antigravity costs nil (unpriced) and deduplicate by responseId - parse date-only CSV rows with UTC calendar - thread fallback calendar through loaders * fix(lint): repair 3113 build and format - calendar before now in makeDailyReport - implicit optional init - wrap long lines and andOperator * style: swiftformat wrap for 3113 * Fix 3113 provider gatekeeper anchors * fix(spend): address 3113 review findings -- freshness, calendar, date-only, fixture model - Preserve cache freshness: return nil when filtered window is empty instead of publishing established zero with now timestamp - Pass pinned calendar into tokenSnapshot for Cursor/Antigravity local snapshots - Keep date-only Cursor CSV rows in configured calendar's noon, not UTC noon - Use clearly fictitious test model test-model-antigravity-a * style: fix line length for fixture model * fix(test): update gatekeeper anchors for CostUsageFetcher line drift Allowlist lines 1339->1335 and 1695->1691 after 075eac7 freshness/calendar fixes * docs: credit #3106 #3113 #3119 changelog entries * feat: add CHF display currency (#3149) * test: fix currency fixtures after CHF became supported * fix(codex): tokscale parity for token counts - max cached, clamped, reasoning split, stale (#3120) * Align Codex token parsing with tokscale stale snapshots - skip lightly regressed cumulative snapshots before interleaved latching - take the maximum of cached and cache-read fields in all parsers - cover cache field selection and out-of-order snapshot accounting with focused tests * Refresh Codex parser hash * Parse bare usage rows in Codex rollouts * Fix stale reasoning and fallback cache parity * Fix Codex fallback test fixture line handling * fix(antigravity): repair offline fallback proof and oauth fallback; fix(spend): limit concurrent dashboard fetches to 3 * fix(lint): break long lines in offline fallback proof tests * fix(tests): update gatekeeper anchors for spend dashboard concurrency limit * fix(tests): correct gatekeeper line anchors for concurrent dashboard fix * docs: update appcast for 0.54.1 * chore: open 0.54.2 unreleased changelog section * Stop re-merging the Codex plan-utilization history with itself on every refresh (#3141) `materializeCodexPlanUtilizationHistoryIfNeeded` exists to fold legacy, opaque and unscoped Codex plan-utilization buckets into the canonical account bucket. Its scoped loop also appended the canonical bucket's own histories to `historiesToMerge` — `matchesTargetContinuity` is true for `rawKey == canonicalKey`, and only the removal of the old key was guarded — so `guard !historiesToMerge.isEmpty` never fired once the canonical bucket had any history, and the migration merge ran on every successful provider refresh and every menu open, merging the history with itself. That merge is quadratic: `updatedPlanUtilizationEntries` copied the whole entry array per entry, scanned it linearly for the insertion point, and allocated the same-hour slice. Measured with an optimized standalone reproduction over a real three-month-old history (session 1909 entries, weekly 2239): 20.6 ms of MainActor time per call, scaling ~3.9x per doubling. `planUtilizationMaxSamples` allows 17520 entries per series, so it would keep growing. Two changes: - Track whether a foreign source actually contributed and require that in the guard, so the canonical-only case returns without merging or rewriting anything. Every path where a legacy, opaque or unscoped bucket contributes is untouched; `legacyRawKeysToRemove` is populated only in branches that also set the flag, so no removal is skipped, and `providerBuckets.unscoped` is cleared only inside the branch that sets it. - Make the merge itself near-linear: `updatedPlanUtilizationEntries` mutates the array in place and finds the insertion point with a binary search for the same strict upper bound (with a fast path for the common append), and `mergedPlanUtilizationHistories` accumulates per series and builds each history once. The binary search assumes entries are sorted by `capturedAt`, which every in-app producer guaranteed through `PlanUtilizationSeriesHistory`'s designated initializer — except the synthesized `Codable` decoder, which assigned entries verbatim from JSON. An explicit `init(from:)` now routes decoding through that initializer, so an on-disk history written by an older build or edited by hand cannot smuggle in an unsorted series. The skipped self-merge also incidentally re-canonicalized per-hour peaks on read; that repair belongs at load time, not on every refresh, and is not reintroduced here. The visible effect is that at most one extra real observation per affected hour is kept. Tests: canonical-only history is returned untouched and enqueues no persistence write (the history revision is unchanged); a genuine foreign merge matches an explicit expected result across overlapping hours, out-of-order sources, distinct series and retention trimming; the binary search's upper-bound contract is pinned directly through a DEBUG shim over an array with a run of equal timestamps (a lower bound would return a different index); and decoding a series whose JSON entries are out of order yields a sorted series. Implemented by grok-4.6 (xhigh) via implementation-loop; reviewed hunk by hunk plus an independent deep review that confirmed both equivalences by differential fuzzing (200k sorted cases with no mismatch) and found the decoder gap, fixed in one iterate round. Gatekeeper line anchors for the touched file were re-verified independently. The DEBUG sortedness assertion is checked once per merged series rather than once per inserted entry: a per-entry check is itself O(n) and reintroduced, in debug builds, exactly the quadratic scan this insertion path removes (measured over the real 4160-entry history: a legacy migration took ~1000 ms with the per-entry assertion versus ~15 ms without it). Co-authored-by: Claude Fable 5 <noreply@anthropic.com> * Fix Codex day cost blanked by trace-only priority turns (#3150) Row ownership evidence compared the retained rows against the persisted standard/priority split using the trace database's tier classification. The persisted maps come from the rows' own pricingMode, so a turn the trace reports as priority after its rows were persisted as standard read as a row-ownership mismatch, the rows lost trust, and the day fell back to the aggregate — which returns nil for long-context tiered models, so the whole day's cost disappeared from the menu, the chart and the window total. Judge retention against both classifications and flag only a group that matches neither. A wrongly retained row set still fails both, because the persisted totals are canonical for the file and tier classification never changes how many tokens the rows carry. Co-authored-by: Claude Opus 5 <noreply@anthropic.com> * docs: credit #3141 and #3150 changelog entries * fix(qwen-cloud): restore Brave browser support in cookie import (#3148) * fix(qwen-cloud): restore Brave browser support, narrowed to Chrome+Brave per AGENTS.md Qwen Cloud's cookie import was restricted to [.chrome] only (commit 529cc6c 'Keep Qwen imports Chrome-only'). Brave users hit 'No Qwen Cloud session cookies found in browsers' even when they had a valid Qwen Cloud session in Brave, because their cookies were never probed. This commit restores Brave in the import order, but follows AGENTS.md L48 ('default Chrome-only when possible to avoid other browser prompts; override via browser list when needed'). The override is the minimum necessary: Chrome + Brave. The other Chromium browsers (chromeBeta, edge, arc, firefox, safari) are deliberately omitted to avoid unsolicited Keychain / browser-store access prompts on automatic refreshes from browsers that don't carry a Qwen Cloud session. Brave is kept because it shares the same Chromium Safe Storage format as Chrome and is a common Qwen Cloud authentication target. Also adds docs/qwen-cloud-proof/README.md with the redacted end-to-end proof captured against the live Qwen Cloud API from the user's Mac after granting the modified binary access to 'Brave Safe Storage' in macOS Keychain. * fix(qwen-cloud): recovery message now names Brave alongside Chrome ClawSweeper P2 follow-up on #3148: when the Brave cookie import fails, QwenCloudSettingsError.missingCookie's recovery message still told users to sign in to Chrome and grant access to Chrome Safe Storage. Now that Brave is a supported source, the message must name both browsers and their respective Safe Storage entries, otherwise a Brave-only user would be told to use Chrome and never find the working path. Updates the error description to: 'No Qwen Cloud session cookies found in browsers. Sign in to Qwen Cloud in Chrome or Brave, allow CodexBar to access the corresponding Safe Storage in Keychain Access (Chrome Safe Storage and/or Brave Safe Storage), or paste a manual Cookie header.' Adds focused test coverage: - missing cookie error mentions both supported browsers and their safe storage - missing cookie error appends non-empty details - missing cookie error omits empty details 35/35 Qwen Cloud tests pass (32 prior + 3 new). * Fix OpenRouter completed-day activity query (#3138) * Preserve unknown Grok period usage (#3159) Co-authored-by: anupamchugh <8416306+anupamchugh@users.noreply.github.com> * fix: report non-writable CLI path conflicts (#3153) * fix: prefer successful CLI install status * fix: keep CLI path conflicts visible * fix: report non-writable CLI path conflicts * docs: add CLI conflict behavior proof * docs: add CLI install comparison screenshots * Fix single-quota icon scaling (#3155) * docs: credit #3138 #3148 #3153 #3155 #3159 changelog entries * fix(spend): silent refresh and invalidation coverage (#3106) * fix(spend): bucket calendar for all heatmap dates and full revision hash - SpendActivityDateFormatting.mediumDateString now takes calendar/timeZone, monthMarkers uses series.calendar, tooltips and accessibility use bucket calendar. - selectedDay renormalized on calendar change to keep toggle correct. - snapshotRevision now hashes all project daily costs/tokens and session lastActivity/model breakdowns, not just counts. Fixes ClawSweeper P2 for #3106. * fix(gatekeeper): update anchors and add provider-specific design markers for spend dashboard * Update provider gatekeeper anchors for v0.54 rebase * fix(test): pin claude spend snapshot in observation test * fix(test): seed pinned claude spend publication before first snapshot * fix: resolve remaining conflict markers from gatekeeper rebase * fix(lint): shorten Sakana test lines * fix(spend): restore heatmap calendar property lost in rebase * Extend spend publication test wait * Restore spend gatekeeper anchors after rebase * fix(spend): sync independent snapshot and bucket calendar normalization for 3106 - publishSpendDashboardTokenSnapshotState now calls synchronizeSharedSpendDashboardAfterTokenPublication - heatmap calendar onChange no longer renormalizes selectedDay via stale controller - SpendDashboardController.update now normalizes selectedDay atomically when bucketTimeZoneIdentifier changes - update gatekeeper anchors for shifted lines (1620,1649,1666,1693) * test(spend): cover independent snapshot sync for 3106 Exercise the direct independent publication path added at UsageStore+SpendDashboardTokenCost.swift:181. The prior focused test seeded Claude before observation and then used the regular Codex publisher, which already syncs independently, so removing that line would not fail. Add a post-start Claude snapshot via _setSpendDashboardTokenSnapshotForTesting and assert the shared dashboard debounced sync is scheduled and the publication inputs update. Verified: swiftformat clean, swiftlint --strict clean, swift test --filter SpendDashboardPublicationTests 18 tests passed. * docs: add fresh-bundle proof for 3106 Add redacted menu-icon crop and dashboard snapshot from debug build 2798eec (swift build --target CodexBarCLI, .build/debug/CodexBarCLI dashboard --pretty). The snapshot shows the shared spend controller produces a dashboard with provider rows/windows, confirming the independent-sync and calendar paths are live in the fresh binary. * docs: add menu and Spend dashboard screenshots for 3106 Add redacted screenshots from fresh debug build 45ba984: - 3106-menu-after-fix.png: menu bar extra open, showing provider rows - 3106-settings-after-fix.png: Settings window (general) - 3106-spend-dashboard-after-fix.png: Usage & Spend pane (usageSpend) with heatmap and Overview, confirming the shared controller renders in the fresh bundle. * docs: remove screenshots for 3106 per request Keep only the redacted CLI dashboard snapshot JSON as fresh-bundle proof; screenshots are not needed. * Improve Antigravity retrieval: retired Flash alias and offline fallback (#3119) * fix(antigravity): allow OAuth errors to fallback to offline when local data exists Fix P2 from Codex review on #3119: AntigravityOAuthFetchStrategy.shouldFallback now checks hasOfflineData, so expired credentials do not block offline. * fix(antigravity): unbind offline account, read app-data, bound scans + proof - Offline snapshot now has nil accountEmail (P1) - OfflineStore also counts $HOME/.gemini/antigravity and .../conversations (P2) - SpendDashboardController bounds Codex scans to 3 concurrent (P2) - Add AntigravityOfflineFallbackProofTests covering app-data and nil email * fix: remove broken proof test, keep P1/P2 fixes and shell proof * fix(gatekeeper): update SpendDashboardController anchors after bounding Codex scans * fix: revert bounded Codex scans (keep offline P1/P2), restore gatekeeper * feat(spend): add tokscale-compatible local readers for Cursor and Antigravity (#3113) * feat(spend): add tokscale-compatible local readers for Cursor and Antigravity - Cursor: read ~/.config/tokscale/cursor-cache/usage*.csv (v1/v2/v3) with tokstyle column handling, cacheWrite = with-without, noon UTC for date-only, and CostUsageDailyReport aggregation. (Sources/CodexBarCore/Providers/Cursor/CursorLocalCSVReader.swift:1) - Antigravity: read ~/.config/tokscale/antigravity-cache/sessions/*.jsonl (tokscale JSONL) and stub for ~/.gemini/antigravity-cli/*.db direct SQLite (ProtoReader to follow). Handles session_meta fallback and dedup. (Sources/CodexBarCore/Providers/Antigravity/AntigravityLocalReader.swift:1) - CostUsageFetcher: local fallback before remote for Cursor (offline) and primary for Antigravity (quota-only before), with Provider-specific by design comments for gatekeeper. (Sources/CodexBarCore/CostUsageFetcher.swift:440) - Antigravity descriptor: enable supportsTokenSnapshot for spend dashboard. (Sources/CodexBarCore/Providers/Antigravity/AntigravityProviderDescriptor.swift:51) Reproduced from /tmp/opencodex/src/adapters/cursor/protobuf-events.ts:218 and /tmp/tokscale/crates/tokscale-core/src/sessions/{cursor,antigravity_cli}.rs Phase 1 of opencodex/tokscale plan, offline-first, no auth. * test(readers): cover cursor csv schemas and antigravity cache fallback * fix(test): include antigravity in cost capable dashboard sources * fix(spend): honor CSV total tokens and add Antigravity Linux capability * fix(test): honor cursor CSV total tokens column in aggregation * fix(spend): repair 3113 tokscale readers P1s - catch remote Cursor errors before falling back to local CSV - recompute summaries after window filtering for Cursor and Antigravity - keep Antigravity costs nil (unpriced) and deduplicate by responseId - parse date-only CSV rows with UTC calendar - thread fallback calendar through loaders * fix(lint): repair 3113 build and format - calendar before now in makeDailyReport - implicit optional init - wrap long lines and andOperator * style: swiftformat wrap for 3113 * Fix 3113 provider gatekeeper anchors * fix(spend): address 3113 review findings -- freshness, calendar, date-only, fixture model - Preserve cache freshness: return nil when filtered window is empty instead of publishing established zero with now timestamp - Pass pinned calendar into tokenSnapshot for Cursor/Antigravity local snapshots - Keep date-only Cursor CSV rows in configured calendar's noon, not UTC noon - Use clearly fictitious test model test-model-antigravity-a * style: fix line length for fixture model * fix(test): update gatekeeper anchors for CostUsageFetcher line drift Allowlist lines 1339->1335 and 1695->1691 after 075eac7 freshness/calendar fixes * test: repair gatekeeper anchors and regenerate parser hash on merged tree --------- Co-authored-by: Yuxin-Qiao <2242016570@qq.com> Co-authored-by: Peter Steinberger <steipete@gmail.com> Co-authored-by: Olddonkey <olddonkeyblog@gmail.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> Co-authored-by: Umut Keltek <35880258+umutkeltek@users.noreply.github.com> Co-authored-by: kiranmagic7 <kiranmagic@proton.me> Co-authored-by: Anupam Chugh <anupam.chugh@gmail.com> Co-authored-by: anupamchugh <8416306+anupamchugh@users.noreply.github.com> Co-authored-by: yicone <yicone@gmail.com> Co-authored-by: Akshay Prabhu <12824090+akshayprabhu200@users.noreply.github.com> * docs: credit #3120 changelog entry * test: fix remaining CHF unconvertible fixtures after #3149 * fix: detect ChatGPT-hosted Codex activity (#3163) * fix(antigravity): reuse signed-in agy for quota refresh (#3161) * docs: credit #3161 and #3163 changelog entries * Fix Claude web cookie refresh (#3162) * docs: credit #3162 changelog entry * chore: regenerate parser hash on merged tree * test: include 0.54.2 parity hash in predecessor list --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> Co-authored-by: Peter Steinberger <steipete@gmail.com> Co-authored-by: Umut Keltek <35880258+umutkeltek@users.noreply.github.com> Co-authored-by: kiranmagic7 <kiranmagic@proton.me> Co-authored-by: Anupam Chugh <anupam.chugh@gmail.com> Co-authored-by: anupamchugh <8416306+anupamchugh@users.noreply.github.com> Co-authored-by: yicone <yicone@gmail.com> Co-authored-by: Akshay Prabhu <12824090+akshayprabhu200@users.noreply.github.com> Co-authored-by: Yuxin Qiao <104957188+Yuxin-Qiao@users.noreply.github.com> Co-authored-by: Yuxin-Qiao <2242016570@qq.com> Co-authored-by: Zihao Qi <35388022+Zihao-Qi@users.noreply.github.com>
This was referenced Aug 24, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Phase 1 of the opencodex/tokscale plan, offline-first, no auth.
Cursor
~/.config/tokscale/cursor-cache/usage*.csv(v1/v2/v3) exactly astokscale/crates/tokscale-core/src/sessions/cursor.rs:182– column indices,cacheWrite = with - without:252,parseCost:132, noon UTC for date-only330, quote-awareparseCSVLine:278.CursorLocalCSVReader.swift:1aggregates toCostUsageDailyReportandCostUsageFetcher.loadCursorLocalSnapshotis tried after the remote dashboard API, so offline or cookie-expired still shows spend.Antigravity
~/.config/tokscale/antigravity-cache/sessions/*.jsonl(tokscaleantigravity syncoutput) astype: usage {modelId, input, output, cacheRead, cacheWrite, timestamp, responseId}sessions/antigravity.rs:48, withsession_metafallback.AntigravityLocalReader.swift:1also stubs direct SQLite~/.gemini/antigravity-cli/conversations/*.dbviaProtoReaderfield numbersantigravity_cli.rs:21for Phase 2.AntigravityProviderDescriptor:51nowsupportsTokenSnapshot:trueso the spend dashboard includes it.Fetcher
CostUsageFetcher.swift:440tries local cache before falling through to the scanner, with// Provider-specific by designfor gatekeeper. Mirrorsopencodex'sOcxUsagecanonical andtokscale's offline-first.Reproduced
git clone --depth1to/tmp/opencodexand/tmp/tokscale, inspectedsrc/adapters/cursor/protobuf-events.ts:218,sessions/{cursor,antigravity_cli}.rs,pricing/aliases.rs:43.Verified:
swift build --target CodexBarCoreok,swiftformatdone, gatekeeper for newprovider == .cursor/.antigravityclusters to be added in follow-up.Branch rebuilt from main (heads
1d2988103→62341aca1)The previously pushed head accidentally deleted the application tree (only
Scripts/survived, ~769k deletions). The branch is now rebuilt off currentmainwith the full feature commit (cherry-picked intact from local history, applies clean) plus:test(readers):CursorAntigravityLocalReaderTestscovering v1/v2/v3 CSV schemas (column indices,cacheWrite = with − without, date-only noon), day aggregation + summary, and the Antigravity JSONL cache withsession_metamodel fallback (5 tests).CostUsageFetcheranchors and added.antigravityto the cost-capable/snapshot provider sets.Real behavior proof
Real behavior proof (after aa464e6)
Remote fallback (P1)
Expired
cursorcookie nowcatchesloadRemoteTokenSnapshotthrow and returnsloadCursorLocalSnapshotif present;antigravityoffline path also covered.Window filtering (P1)
30-day view now recomputes
summaryfromfiltered(wasfull.summarywith all-time totals).Unpriced Antigravity (P1)
AntigravityLocalReadernow leavescostUSD/totalCostUSDnilandcostProvenance.unknown(was0+listPriceEstimate→ $0).Dedup / UTC / Calendar (P1/P2)
Real behavior proof — PR #3113 (terminal only, no screenshots)
Synthetic
TOKSCALE_CONFIG_DIRcaches were created locally (no real cookies, no secrets) and read through the verbatimCursorLocalCSVReader/AntigravityLocalReaderat902c7401a(filesSources/CodexBarCore/Providers/Cursor/CursorLocalCSVReader.swiftandSources/CodexBarCore/Providers/Antigravity/AntigravityLocalReader.swift).Covers:
cacheWrite = with - without, noon UTC for date-only, quote-awareparseCSVLine)Total Tokenscolumn honored as authoritativetype: usagewithsession_metamodel fallback,responseIddedup, and unpriced (costUSD nil) handlingPrepared from:
TOKSCALE_CONFIG_DIR=/tmp/tokscale-proofwith synthetic CSV/JSONL, compiled viaswiftc -module-cache-path /tmp/clang-cache /tmp/proof-3113-main.swift -o /tmp/proof-3113-binand executed asTOKSCALE_CONFIG_DIR=/tmp/tokscale-proof /tmp/proof-3113-bin.Fix 075eac7 — address 4 review blockers + production fetcher proof
Fixes pushed at 075eac7:
CostUsageFetcher.swift:1190Preserve freshness:loadCursorLocalSnapshot/loadAntigravityLocalSnapshotnowguard !filtered.isEmpty else { return nil }instead of publishing an established zero withnowwhen the window has no rows (stale cache no longer looks fresh)CostUsageFetcher.swift:1210Pass pinned calendar:tokenSnapshot(..., calendar: cal, ...)so the current-day session lookup uses the configured calendar, notCalendar.currentCursorLocalCSVReader.swift:184Keep date-only rows in configured day: parseyyyy-MM-ddas year/month/day in UTC, then reconstruct at 12:00 in the passedcalendar(UTC+13/14 no longer shifts to next local day)CursorAntigravityLocalReaderTests.swift:111Use fictitious fixturetest-model-antigravity-ainstead ofgemini-3.7-flashProduction fetcher proof —
CostUsageFetcher.loadCursorLocalSnapshot/loadAntigravityLocalSnapshot(fixed 075eac7, syntheticTOKSCALE_CONFIG_DIR=/tmp/tokscale-proof):Fix 60fdee7 — lint
test-model-ato fixline_lengthviolation atCursorAntigravityLocalReaderTests.swift:112(was 126 chars, now <120).swiftlint --strictnow passes (0 violations). Previous proofs'test-model-a/test-gemini-modelare both clearly fictitious and acceptable; updated synthetic cache now usestest-model-aconsistently, see combined proof above (reader + fetcher).Head is now
60fdee7, CI re-running.