docs(changelog): collect spend refresh and Claude cost improvements

[steipete]

Collect the batch's user-visible notes in one independent changelog PR, preserving all existing Unreleased entries and Peter's concurrent Overview fix #3482. Lead with spend reliability, then performance and the bounded opt-in CLI addition. Released sections are unchanged.

Merge this after the five code PRs:

• fix(spend): refresh expired dashboard snapshots on return #3107: refresh expired Usage & Spend charts and handle midnight rollover.
• perf: persist Claude cost report memo across launches #3284: reuse compatible Claude local-cost reports across launches.
• feat(claude): add opt-in cost breakdown with accurate periods #3244: opt-in Claude CLI daily/model detail with accurate periods and partial attribution.
• fix(codex): resume interrupted warm cost scans #3490: preserve unfinished warm Codex scan work; Spend Dashboard Codex publish delayed for hours — stale scanSinceKey reseeds lookback queue on every scan #3411 remains open for the broader managed-account report.
• fix(antigravity): retain unknown usage across CLI quota lanes #3489: Antigravity CLI quota-bucket parity with unknown/disabled availability preserved.

Validation: local and final committed-branch autoreview clean through P2; checked that all five entries and #3482 are present and that released sections are byte-for-byte unchanged. Runtime proof and full suites are recorded on the code PRs. Exact-head notes CI succeeded on 69c93cd032fbf5582f972bf26d6923986417c8a1; all applicable checks are green.

This prepares notes only. The remaining queue still contains autonomous work, so there is no release/tag/publication in this batch. The existing pending version remains 0.56.9.

[clawsweeper]

🦞👀
ClawSweeper picked this up.

Pull request received. I will update this pull request when review starts.

ClawSweeper review complete

ClawSweeper finished reviewing this revision. The review result is being finalized.

View the workflow run.

[clawsweeper]

Codex review: blocked before merge. Reviewed September 8, 2026, 12:43 AM ET / 04:43 UTC.

ClawSweeper review

What this changes

Adds Unreleased notes for spend refresh, resumable Codex scans, Claude cost caching and optional breakdowns, and Antigravity quota display.

Merge readiness

⛔ Blocked before merge - 2 items remain

Keep open: this owner-authored changelog PR remains useful, and the reviewed diff has no actionable defect. Its five implementation prerequisites remain open in the captured context.

Priority: P3
Reviewed head: 69c93cd032fbf5582f972bf26d6923986417c8a1

Review scores

Measure	Result	What it means
Overall readiness	🐚 platinum hermit (4/6)	A focused, correctly scoped changelog update with verified preservation of existing notes.
Proof confidence	🌊 off-meta tidepool	Not applicable: This owner-authored PR changes only changelog prose; runtime evidence belongs to the prerequisite implementation PRs.
Patch quality	🐚 platinum hermit (4/6)	No actionable review findings were identified.

Verification

Check	Result	Evidence
Real behavior	Not applicable	Not applicable: This owner-authored PR changes only changelog prose; runtime evidence belongs to the prerequisite implementation PRs.
Evidence reviewed	8 items	Complete introduced diff: The independently inspected base-to-head diff changes only CHANGELOG.md, adding five scoped entries and a highlight while moving the existing Performance section. Preservation verified: A read-only Python comparison of the base and head Git blobs confirmed that all 16 prior Unreleased bullets remain and the entire released portion beginning at 0.56.8 is byte-for-byte identical. Explicit owner sequencing: The owner’s comment requires landing #3107, #3284, #3244, #3490, and #3489 first and matching their final scope. All five are open in the supplied snapshot. The broader #3411 remains open. Source: #3491 (comment).
Findings	None	None.
Security	None	None.

How this fits together

CodexBar’s changelog collects user-facing changes for upcoming releases. Release tooling extracts its version section for GitHub release notes and Sparkle update notes.

flowchart LR
  A[Implementation PRs] --> B[Confirmed merged scope]
  B --> C[Unreleased changelog]
  C --> D[Release preparation]
  D --> E[GitHub release notes]
  D --> F[Sparkle update notes]

Loading

Before merge

• Resolve merge risk (P1) - The prerequisite PRs’ final merged scope may differ from these prepared notes.
• Complete next step (P2) - Land fix(spend): refresh expired dashboard snapshots on return #3107, perf: persist Claude cost report memo across launches #3284, feat(claude): add opt-in cost breakdown with accurate periods #3244, fix(codex): resume interrupted warm cost scans #3490, and fix(antigravity): retain unknown usage across CLI quota lanes #3489 first, then confirm these notes match their final merged scope.

Agent review details

Security

None.

Review metrics

Metric	Value	Why it matters
Changelog preservation	16 existing bullets retained; released sections unchanged	The batch notes preserve concurrent release-note work and published history.

Merge-risk options

Maintainer options:

1. Decide the mitigation before merge
   Keep one consolidated Unreleased section that accurately reflects the landed batch and preserves the broader managed-account report’s unresolved scope.
2. Pause or close
   Do not merge this PR until maintainers decide whether the risk is worth taking.

Technical review

Best possible solution:

Keep one consolidated Unreleased section that accurately reflects the landed batch and preserves the broader managed-account report’s unresolved scope.

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

Not applicable: this PR changes release-note prose; direct diff and blob comparisons establish its scope and preservation properties.

Is this the best way to solve the issue?

Yes: consolidating the batch’s notes is a focused approach consistent with the owner’s explicit sequencing and existing release process.

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning medium; reviewed against 2a71b479a5d4.

Labels

Label justifications:

• P3: This is bounded release-note preparation with no production behavior change.
• rating: 🐚 platinum hermit: Overall readiness is 🐚 platinum hermit; proof is 🌊 off-meta tidepool and patch quality is 🐚 platinum hermit.
• status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR. Not applicable: This owner-authored PR changes only changelog prose; runtime evidence belongs to the prerequisite implementation PRs.

Evidence

What I checked:

• Complete introduced diff: The independently inspected base-to-head diff changes only CHANGELOG.md, adding five scoped entries and a highlight while moving the existing Performance section. (CHANGELOG.md:3, 69c93cd032fb)
• Preservation verified: A read-only Python comparison of the base and head Git blobs confirmed that all 16 prior Unreleased bullets remain and the entire released portion beginning at 0.56.8 is byte-for-byte identical. (CHANGELOG.md:37, 69c93cd032fb)
• Explicit owner sequencing: The owner’s comment requires landing fix(spend): refresh expired dashboard snapshots on return #3107, perf: persist Claude cost report memo across launches #3284, feat(claude): add opt-in cost breakdown with accurate periods #3244, fix(codex): resume interrupted warm cost scans #3490, and fix(antigravity): retain unknown usage across CLI quota lanes #3489 first and matching their final scope. All five are open in the supplied snapshot. The broader Spend Dashboard Codex publish delayed for hours — stale scanSinceKey reseeds lookback queue on every scan #3411 remains open. Source: docs(changelog): collect spend refresh and Claude cost improvements #3491 (comment). (69c93cd032fb)
• Release-note consumer: The release documentation says GitHub and Sparkle notes are extracted from the changelog; release automation rejects an Unreleased top section. This PR leaves the pending version at 0.56.9. (docs/RELEASING.md:21, 69c93cd032fb)
• Current-main routing history: Available main history records Peter Steinberger on recent changelog consolidation and the merged Overview update. Full blame and parent-blob inspection failed because missing objects could not be fetched; no source-line introduction attribution is asserted. (CHANGELOG.md, 2a71b479a5d4)
• Release baseline: The local v0.56.8 tag resolves to the recorded appcast update. The proposed entries belong to the subsequent Unreleased section, not an already shipped section. (CHANGELOG.md:3, 6ef82690b4a7)

Likely related people:

• steipete: Suggested for follow-up; no historical authorship or introduction is verified. (role: unverified routing candidate; confidence: low)

Rating scale

Score	Internal tier	Crab rank	Meaning
6/6	S	🦀 challenger crab	Exceptional readiness
5/6	A	🦞 diamond lobster	Very strong readiness
4/6	B	🐚 platinum hermit	Good normal PR; ordinary maintainer review
3/6	C	🦐 gold shrimp	Useful, but confidence is limited
2/6	D	🦪 silver shellfish	Proof or implementation needs work
1/6	F	🧂 unranked krab	Not merge-ready
N/A	NA	🌊 off-meta tidepool	Rating does not apply

Overall follows the weaker of proof and patch quality.
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.

Workflow

• 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.

History

Review history (2 earlier review cycles)

• reviewed 2026-09-08T03:49:31.004Z sha 4ef2014 :: blocked before merge. :: none
• reviewed 2026-09-08T04:12:38.454Z sha 69c93cd :: blocked before merge. :: none

[steipete]

Notes verification complete on 69c93cd032fbf5582f972bf26d6923986417c8a1.

Local and final committed-branch autoreview are clean through P2. The diff changes only CHANGELOG.md; all five prepared entries, prior Unreleased bullets, contributor credit, and the concurrent Overview fix remain present. Released sections are byte-for-byte unchanged. CI run 34185810334 succeeded on this exact head; all applicable checks are green.

Merge after #3107, #3284, #3244, #3490, and #3489, and retain their final scope in these notes. Their runtime and full-suite proof belongs to those code PRs. The broader #3411 report stays open. This is release-note preparation only; no tag, release, or publication was performed.