[codex] Surface Claude CLI scoped weekly limits

[janpollak]

• Preserve model-scoped weekly limits from the Claude CLI /usage parser as named extra windows.
• This surfaces Current week (Fable) from Claude CLI as Fable only in CodexBar, matching the OAuth/Web scoped-limit behavior.
• Handles live CLI reset variants (Reset vs Resets) and falls back to the all-model weekly reset when a scoped weekly reset line is omitted by the TUI capture.

Verification:

• swift test --filter ClaudeUsageTests
• Live proof: .build/arm64-apple-macosx/debug/CodexBarCLI usage --provider claude --source cli --format json --pretty
  • Claude CLI 2.1.204
  • extraRateWindows[0].id: claude-weekly-scoped-fable
  • extraRateWindows[0].title: Fable only
  • extraRateWindows[0].window.usedPercent: 70
  • extraRateWindows[0].window.resetDescription: Resets Jul 10 at 3am (Europe/Prague)
  • extraRateWindows[0].window.resetsAt: 2026-07-10T01:00:00Z

[clawsweeper]

Codex review: needs maintainer review before merge. Reviewed July 8, 2026, 12:56 PM ET / 16:56 UTC.

Summary
The branch parses Claude CLI model-scoped Current week (...) panes into named extra rate windows, wires them into Claude CLI snapshots, merges them with web extras, and adds focused Swift tests.

Reproducibility: not applicable. as a strict bug reproduction: this is a provider-data display improvement. Source inspection shows current main drops CLI extra windows, while the PR provides focused parser/fetcher coverage and copied live CLI output for the new path.

Review metrics: 3 noteworthy metrics.

• Diff scope: 3 files, +236/-7. The change is bounded to Claude CLI parsing, snapshot wiring, and one focused test file.
• Review continuity: 1 prior P2 finding addressed. The latest head adds CLI/web extra-window merge and dedupe logic for the earlier web-extra replacement concern.
• Validation signal: 1 focused test command claimed, 1 live CLI output proof. The PR body reports swift test --filter ClaudeUsageTests and copied live CLI output showing the new scoped Fable window.

Root-cause cluster
Relationship: canonical
Canonical: #1993
Summary: This PR is the active CLI-specific implementation for Claude scoped weekly windows; the merged OAuth/Web scoped-window PR is related but did not cover the CLI TUI parser path.

Members:

• partial_overlap: Claude: surface per-model weekly quota carve-outs (e.g. Fable 5) from limits[] #1851 - The merged PR introduced OAuth/Web scoped weekly limits and the shared extra-window mapper, but current main still drops CLI-scoped windows.

Proposal only: this assessment does not dispatch repair, suppress jobs, mutate sibling items, close, or merge anything.

Merge readiness
Overall: 🐚 platinum hermit
Proof: 🐚 platinum hermit
Patch quality: 🐚 platinum hermit
Result: ready for maintainer review.

Overall follows the weaker of proof and patch quality, so missing proof can cap an otherwise strong patch.

Rank-up moves:

• none.

Next step before merge

• [P2] No repair lane is needed; the remaining action is ordinary maintainer review of a focused, proven PR.

Security
Cleared: The diff changes Swift parsing/wiring and tests only; no dependency, workflow, secret, release, or supply-chain execution surface changed.

Review details

Best possible solution:

Land the focused CLI parser and wiring after maintainer review, keeping CLI scoped-window IDs, titles, and web-extra merging aligned with the existing Claude scoped-weekly behavior.

Do we have a high-confidence way to reproduce the issue?

Not applicable as a strict bug reproduction: this is a provider-data display improvement. Source inspection shows current main drops CLI extra windows, while the PR provides focused parser/fetcher coverage and copied live CLI output for the new path.

Is this the best way to solve the issue?

Yes: extraRateWindows is already the Claude rendering surface for scoped limits, and the latest diff keeps the CLI implementation narrow while preserving web extras. I do not see a smaller supported path that would cover CLI TUI scoped panes.

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning high; reviewed against aa401f1d8b74.

Label changes

Label justifications:

• P2: This is a normal-priority Claude provider usage-display improvement with limited parser and snapshot-wiring blast radius.
• rating: 🐚 platinum hermit: Overall readiness is 🐚 platinum hermit; proof is 🐚 platinum hermit and patch quality is 🐚 platinum hermit.
• status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR. Sufficient (live_output): The PR body includes copied live CLI output showing the after-fix Claude scoped Fable extra window with ID, title, percent, reset text, and parsed reset timestamp.
• proof: sufficient: Contributor real behavior proof is sufficient. The PR body includes copied live CLI output showing the after-fix Claude scoped Fable extra window with ID, title, percent, reset text, and parsed reset timestamp.

Evidence reviewed

What I checked:

• Repository policy read: Read the full 48-line AGENTS.md; its guidance to prefer focused CLI/parser tests, avoid unrequested live provider probes, keep provider data siloed, and not rely on Claude CLI status lines was applied to this review. (AGENTS.md:23, aa401f1d8b74)
• Current main drops CLI extras: Current main builds Claude CLI snapshots with an empty extraRateWindows array, so CLI-only scoped weekly panes are not surfaced there. (Sources/CodexBarCore/Providers/Claude/ClaudeUsageFetcher.swift:1349, aa401f1d8b74)
• Existing scoped-window precedent: Current main already maps OAuth/Web weekly_scoped Claude limits into NamedRateWindow, which makes the CLI behavior consistent with an existing rendering path rather than a separate UI concept. (Sources/CodexBarCore/Providers/Claude/ClaudeScopedWeeklyLimitMapper.swift:14, aa401f1d8b74)
• PR parser implementation: The PR extracts non-all-model Current week (...) blocks, skips Sonnet/Opus lanes, and maps remaining scoped weekly usages to named extra windows with parsed reset data. (Sources/CodexBarCore/Providers/Claude/ClaudeStatusProbe.swift:394, 5f6a7cc6edfc)
• PR snapshot and merge wiring: The latest PR head passes CLI snap.extraRateWindows into ClaudeUsageSnapshot and merges/deduplicates CLI and web extra windows, addressing the prior review comment about losing web-only extras. (Sources/CodexBarCore/Providers/Claude/ClaudeUsageFetcher.swift:1349, 5f6a7cc6edfc)
• Focused regression coverage: The added tests cover Fable and Compass scoped CLI panes, singular Reset, fallback reset inheritance, no Opus pollution, and web-extra-window merge/dedupe behavior. (Tests/CodexBarTests/ClaudeUsageTests.swift:50, 5f6a7cc6edfc)

Likely related people:

• steipete: Git blame and file history tie the current Claude CLI parser and snapshot wiring to recent Claude provider work by this contributor. (role: recent area contributor; confidence: high; commits: a83a83fa4131, cfd747f8, 02bd8e8d; files: Sources/CodexBarCore/Providers/Claude/ClaudeStatusProbe.swift, Sources/CodexBarCore/Providers/Claude/ClaudeUsageFetcher.swift)
• konon4: Authored the merged OAuth/Web scoped weekly limit implementation and shared mapper that this CLI path is matching. (role: introduced related behavior; confidence: high; commits: 21e3beb37743; files: Sources/CodexBarCore/Providers/Claude/ClaudeScopedWeeklyLimitMapper.swift, Sources/CodexBarCore/Providers/Claude/ClaudeUsageFetcher.swift, Sources/CodexBarCore/Providers/Claude/ClaudeWeb/ClaudeWebExtraRateWindowParser.swift)
• Yuxin-Qiao: Recent history includes Claude CLI subscription /usage and fallback handling near the same parser/fetcher surface. (role: adjacent CLI fallback contributor; confidence: medium; commits: 126dc923; files: Sources/CodexBarCore/Providers/Claude/ClaudeStatusProbe.swift, Sources/CodexBarCore/Providers/Claude/ClaudeUsageFetcher.swift)

What the crustacean ranks mean

• 🦀 challenger crab: rare, exceptional readiness with strong proof, clean implementation, and convincing validation.
• 🦞 diamond lobster: very strong readiness with only minor maintainer review expected.
• 🐚 platinum hermit: good normal PR, likely mergeable with ordinary maintainer review.
• 🦐 gold shrimp: useful signal, but proof or patch confidence is still limited.
• 🦪 silver shellfish: thin signal; proof, validation, or implementation needs work.
• 🧂 unranked krab: not merge-ready because proof is missing/unusable or there are serious correctness or safety concerns.
• 🌊 off-meta tidepool: rating does not apply to this item.

Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics.

How this review workflow works

• ClawSweeper keeps one durable marker-backed review comment per issue or PR.
• Re-runs edit this comment so the latest verdict, findings, and automation markers stay together instead of adding duplicate bot comments.
• A fresh review can be triggered by eligible @clawsweeper re-review comments, exact-item GitHub events, scheduled/background review runs, or manual workflow dispatch.
• PR/issue authors and users with repository write access can comment @clawsweeper re-review or @clawsweeper re-run on an open PR or issue to request a fresh review only.
• Maintainers can also comment @clawsweeper review to request a fresh review only.
• Fresh-review commands do not start repair, autofix, rebase, CI repair, or automerge.
• Maintainer-only repair and merge flows require explicit commands such as @clawsweeper autofix, @clawsweeper automerge, @clawsweeper fix ci, or @clawsweeper address review.
• Maintainers can comment @clawsweeper explain to ask for more context, or @clawsweeper stop to stop active automation.

Review history (5 earlier review cycles)

• reviewed 2026-07-08T08:03:29.531Z sha a50ebd4 :: needs real behavior proof before merge. :: none
• reviewed 2026-07-08T08:11:49.070Z sha a50ebd4 :: needs real behavior proof before merge. :: none
• reviewed 2026-07-08T09:37:56.057Z sha b174473 :: needs maintainer review before merge. :: none
• reviewed 2026-07-08T09:44:07.408Z sha b174473 :: needs maintainer review before merge. :: none
• reviewed 2026-07-08T13:37:29.149Z sha b174473 :: needs maintainer review before merge. :: none

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: b174473925

ℹ️ About Codex in GitHub

Codex has been enabled to automatically 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 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".