Preserve Cursor model usage above 100%

[HemSoft]

Closes #187

Summary

• preserve raw finite Cursor Models and Other Models percentages above 100
• keep negative bucket values at zero and omit values that cannot be safely rounded for display
• cover dashboard text, capped gauge fill, projection input, and History persistence with an over-limit parser regression

Verification

• git diff --check
• ./scripts/test-cut-changelog.sh
• ./scripts/test-release-artifacts.sh
• ./scripts/test-run.sh
• ./test.sh reached Xcode package resolution, then repeated device-enumeration attempts against a connected passcode-locked device; interrupted before compilation with no test failure
• direct Debug xcodebuild ... build attempt hit the same passcode-locked device enumeration loop and was interrupted before compilation

Residual risk

The local XCTest and build steps could not start while Xcode repeatedly queried the locked connected device. GitHub Actions remains the independent build-and-test gate.

---

Summary by cubic

Fixes #187 by preserving Cursor Models and Other Models usage above 100% in dashboard text, projection inputs, and History values, while keeping gauge fill capped at full.

• Negative values are still clamped to zero, and values that can't be safely rounded are omitted.
• Adds regression coverage for over-limit values, capped gauges, projections, History, and the Mac subtitle.

^{Written for commit 972f30d. Summary will update on new commits.}

Greptile Summary

The PR preserves finite Cursor model-usage percentages above 100% across dashboard text, projections, and history while retaining capped gauge fills and rejecting values that cannot be safely displayed.

• Removes the 100% cap from stored and formatted Cursor model-usage values.
• Validates that decoded percentages can be safely rounded to an integer.
• Adds regression coverage for over-limit, negative, and excessively large percentages.
• Documents the corrected behavior in the changelog.

Confidence Score: 5/5

The PR appears safe to merge.

No blocking failure remains.

Important Files Changed

Filename	Overview
CodexBarMac/Services/CursorUsageProvider.swift	Preserves representable usage percentages above 100%, keeps negative values at zero, and omits values that cannot safely be converted for display.
CodexBarMacTests/APIKeyAndCursorProviderTests.swift	Covers over-limit values through display, gauge, projection, and history paths, plus negative and non-representable inputs.
CHANGELOG.md	Documents preservation of Cursor model usage above 100% and continued gauge capping.

_{Reviews (2): Last reviewed commit: "Assert Cursor details through Mac subtit..." | Re-trigger Greptile}

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-02T07:30:55.738809Z	972f30d	Manual request

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" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[HemSoft]

@codex review

Dismissed because a newer commit was pushed; Greptile will re-review the current head.

[HemSoft]

@codex review

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. What shall we delve into next?

Reviewed commit: 972f30d907

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".

[HemSoft]

Readiness audit for current head 972f30d9075efa399851422d81c83e6f43bc8b16: required Build and Test passed 441 tests with 0 failures; Greptile approved this exact head and its review check passed; Cubic completed successfully on this exact head; Codex completed a clean exact-head review; and the GraphQL audit found zero review threads. Local git diff --check and all three smoke scripts passed. Local Xcode build/test attempts were blocked before compilation by repeated enumeration of a connected passcode-locked device, while CI compiled and tested the same head successfully. GitHub reports CLEAN, MERGEABLE, and APPROVED. PR remains open for the parent merge gate.