fix(shared): keep the newest shared usage scan

[yashranaway]

What Changed

Choose the most recently read usage summary for each shared transcript directory. Use environment IDs to break ties so selection remains stable.

Why

When two environments read the same directory, the merge always keeps the environment with the smallest ID. An older cached response can therefore hide tokens and costs present in a newer response, including while the older environment is failing to refresh.

The shared merge serves web, desktop, and mobile.

Testing

• usageMerge.test.ts passes, 13/13.
• The regression returns the stale $4 total before the fix and the current $10 total afterward, regardless of response order.
• Shared package typecheck, targeted lint, and formatting pass.

Checklist

• One focused change
• Explained what changed and why
• Added regression coverage for the changed behavior
• No UI layout or animation changes

Model: GPT-6 Astra
Harness: T3 code

Note

Fix claimSources to keep newest shared usage scan by readAt

When environments report the same physical transcript directory, mergeUsage now keeps the provider buckets from the environment with the newest summary readAt time instead of using environment-id order.

• Changes source ownership ordering in usageMerge.ts to descending parsed readAt, with environment id as tie-breaker; invalid or absent dates fall back to zero for ordering.
• Adds a regression test in usageMerge.test.ts verifying the newer environment supplies merged totals regardless of input order.
• Risk: environments with missing or unparseable readAt dates order as zero and may lose ownership to any environment with a valid date.

^{Macroscope summarized 78eded7.}

Summary by CodeRabbit

• Bug Fixes
  • Usage data from duplicate environment scans now consistently prioritizes the most recently read summary.
  • Duplicate-source reporting is now deterministic, regardless of input order.
  • Ties are resolved consistently using the environment identifier.

[macroscopeapp]

Approvability

Verdict: Approved at 78eded7

Macroscope's review found this PR approvable — This is a focused correction to shared usage deduplication: duplicate transcript directories now retain the newest scan while preserving deterministic behavior and existing non-duplicate totals. The accompanying regression test covers both input orders, with no schema, deployment, or configuration changes.

^{You can add or adjust custom eligibility rules. Learn more.}

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Advanced

Run ID: 85224724-c2e6-481b-8ac3-f2c3bb311236

📥 Commits

Reviewing files that changed from the base of the PR and between 1ab2dfb and a51235a.

📒 Files selected for processing (2)

• packages/shared/src/usageMerge.test.ts
• packages/shared/src/usageMerge.ts

Included review availability: Your plan provides up to 10 included reviews per hour; 9 remain after this review.

---

📝 Walkthrough

Walkthrough

The merge logic now assigns shared transcript sources to the environment with the newest readAt. A test verifies the selected usage data and duplicate reporting remain identical when input order changes.

Changes

Usage merge selection

Layer / File(s)	Summary
Newest summary winner and validation packages/shared/src/usageMerge.ts, packages/shared/src/usageMerge.test.ts	claimSources sorts environments by descending summary.readAt, with environment ids as tie-breakers. Tests verify the newest environment wins and that results are independent of input order.

Priority: ⬇️ Low

Estimated code review effort: 2 (Simple) | ~10 minutes

Change: Bug fix

Suggested reviewers: maria-rcks

Merge Risk: ⚪ Minimal · up to a5123

The usage winner-selection change is supported by order-independent regression coverage and is mergeable.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and concisely describes the main change: retaining the newest shared usage scan.
Description check	✅ Passed	The description explains what changed, why it changed, how it was tested, and confirms that the focused-change checklist is complete. UI changes are addressed as not applicable.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 1 functions across 2 files.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create a new PR

---

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