revert(web): restore sparse hourly usage breakdown

[t3-code]

What Changed

• reverts the hourly breakdown behavior introduced by fix(web): usage hourly breakdown lists every hour chronologically #7595
• restores sparse, newest-first hourly rows instead of inserting 24 chronological rows
• adds a component regression test proving recent non-zero values stay visible first

Why

The zero-filled chronological table put the oldest hours first, so recent non-zero usage values could be buried below a screen of empty rows and appear to have disappeared. The chart remains unchanged and still shows the complete 24-hour window.

UI Changes

Browser screenshots were skipped at the requester's direction. Before this fix, Hour rendered 24 rows from oldest to newest. After this fix, it renders only reported activity from newest to oldest.

Verification

• regression test failed on main with 24 rows instead of 2, then passed after the revert
• 17 focused usage tests pass
• web typecheck passes
• changed-file lint and formatting pass
• git diff --check passes

Built with GPT-5.6-sol in the Codex harness.

request provenance

• requested by: @maria-rcks

---

Note

Low Risk
UI-only table ordering/filtering for usage breakdown; no auth, data, or API changes. The chart still uses the full hourly window.

Overview
The 24h usage Hour breakdown no longer pads empty hours or lists oldest-first. It now shows only reported hourly rows, newest first, so recent spend is not buried under a screen of zeros. The chart still covers the full window.

Adds a UsagePage regression test that the table has two activity rows with the newer $11.00 above $13.00.

^{Reviewed by Cursor Bugbot for commit dcc98ac. Bugbot is set up for automated code reviews on this repo. Configure here.}

Note

Restore sparse hourly usage breakdown in UsagePage

Stops synthesizing zero-filled empty hours for 1-day windows in UsagePage.tsx. The breakdown now shows only hours present in merged.hourly, ordered newest-first, and the zeroHour helper is removed. Adds a test suite in UsagePage.test.tsx that verifies sparse rows and newest-first ordering.

^{Macroscope summarized dcc98ac.}

[macroscopeapp]

Approvability

Verdict: Approved at dcc98ac

Macroscope's review found this PR approvable — This revert simplifies hourly usage table display by showing only hours with actual data instead of zero-filling every hour. The change removes complexity, adds a test, and restores previous behavior with clear intent.

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

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	13.4 KiB	13.5 KiB	+33 B (+0.2%)	15.1 KiB	✅
Codex	Thread snapshot wire	6.9 KiB	6.9 KiB	+13 B (+0.2%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.5 KiB	6.6 KiB	+20 B (+0.3%)	7.8 KiB	✅
Codex	Live turn WebSocket decoded	55.0 KiB	55.0 KiB	0 B (0.0%)	66.4 KiB	✅
Codex	Live turn messages	16	16	0 (0.0%)	21	✅
Claude	Total thread wire	13.5 KiB	13.4 KiB	−29 B (−0.2%)	15.1 KiB	✅
Claude	Thread snapshot wire	6.9 KiB	6.9 KiB	+1 B (+0.0%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.6 KiB	6.5 KiB	−30 B (−0.4%)	7.8 KiB	✅
Claude	Live turn WebSocket decoded	55.8 KiB	55.8 KiB	0 B (0.0%)	66.4 KiB	✅
Claude	Live turn messages	16	16	0 (0.0%)	21	✅

Baseline: 21e80a0 · PR result: dcc98ac · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 109.4 KiB
• Claude decoded thread snapshot: 110.1 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.