perf(mobile): skip impossible thread-key lookups

[juliusmarminge]

Home builds scoped thread-key strings while checking an empty queued-message set and searches off-page settled rows even when no thread is selected. With a large settled history, streaming updates repeat this unnecessary work across thousands of rows.

Treat an empty queued-key set as absent in the three list helpers, so optional chaining skips key construction. Skip the settled-tail slice/search when the selected key is null. Nonempty queues and off-page selections keep their existing behavior. No cache, new state or sorting algorithm is added.

Release measurements

Two alternating before/after pairs on iPhone 17 Pro, iOS 27 simulator. Home visible with 1,210 shells; five streams at 20 Hz for 15 seconds, 1,500 emitted events each run. The same server snapshot and app cache are restored per run. Same optimized native Release executable, Hermes -O, dev=false, no Metro server. Source-map comparison confirms only threadListV2.ts differs. Baseline 510e5bb, candidate 4ff5c69.

Measurement	Before 1	After 1	Before 2	After 2
List-helper inclusive samples	496	366	549	389
Active JS samples	2,417	2,058	2,571	2,474
Total samples	15,591	15,023	15,094	14,995
GC samples	298	223	300	282
Process CPU seconds	4.67	3.90	4.67	4.54
Capture seconds	24.03	22.91	22.87	22.88

Both pairs reduced CPU, but the magnitude varied. The first baseline capture was 1.1 seconds longer; the second pair had matching capture lengths and reduced CPU by 0.13 seconds. These two pairs do not establish a stable percentage improvement. The list-helper count is the union of samples containing any threadListV2.ts frame, including remaining work. Sample counts are not milliseconds, and process CPU includes native threads and profiler overhead. No whole-app latency or battery claim.

evidence.json · empty-key-after-1.cpuprofile · empty-key-after-2.cpuprofile · empty-key-before-1.cpuprofile · empty-key-before-2.cpuprofile

Verification

106 focused list and move-availability tests pass, including nonempty queues, empty queues and selected settled rows beyond the visible page. Targeted lint and mobile typechecking pass. Exact set membership and non-null selection logic are unchanged.

This changes shared iOS/Android list helpers used by Home, the tablet sidebar and arrangement UI. Runtime measurement was iOS only. Web, desktop, providers and contracts are unchanged. Unrelated widget edits are excluded. No visible UI change is intended.

Julius requested one PR per measured hotspot in this maintainer mobile performance audit.

Implementation: Claude Opus 5.5 through T3 Code. Integration, review and release profiling: GPT-6 Astra through the Codex harness.

---

[macroscopeapp]

Approvability

Verdict: Approved at 4ff5c69

Macroscope's review found this PR approvable — This is a focused one-file mobile performance optimization that removes impossible string allocations and settled-row searches while preserving list membership and selection behavior. It introduces no product-default, schema, infrastructure, or user-facing changes.

No code changes detected at f0707ab. Prior analysis still applies.

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

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ No successful main baseline artifact is available yet. This run establishes the initial measurement.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	—	5.0 KiB	—	6.8 KiB	✅
Codex	Thread snapshot wire	—	3.8 KiB	—	4.9 KiB	✅
Codex	Live turn WebSocket wire	—	1.2 KiB	—	2.0 KiB	✅
Codex	Live turn WebSocket decoded	—	20.9 KiB	—	29.3 KiB	✅
Codex	Live turn messages	—	2	—	8	✅
Claude	Total thread wire	—	5.0 KiB	—	6.8 KiB	✅
Claude	Thread snapshot wire	—	3.8 KiB	—	4.9 KiB	✅
Claude	Live turn WebSocket wire	—	1.2 KiB	—	2.0 KiB	✅
Claude	Live turn WebSocket decoded	—	21.2 KiB	—	29.3 KiB	✅
Claude	Live turn messages	—	1	—	8	✅

Baseline: unavailable · PR result: f0707ab · 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: 108.5 KiB
• Claude decoded thread snapshot: 108.8 KiB

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

[coderabbitai]

Important

Review skipped

We couldn't safely recover the incremental review. No full review was started, and the last reviewed checkpoint was preserved. Retry later, or explicitly request a full review by commenting @coderabbitai full review.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

• Configuration used: Repository: pingdotgg/t3code/.coderabbit.yaml
• Review profile: CHILL
• Plan: Team
• Run ID: 5563d80c-dbe0-4a03-a2ea-102508ddd01e

📥 Commits

Reviewing files that changed from the base of the PR and between 510e5bb and 4ff5c69.

📒 Files selected for processing (1)

• apps/mobile/src/features/threads/threadListV2.ts

Limit details: You’ve used all 10 included reviews currently available.

---

📝 Walkthrough

Walkthrough

The thread list now treats empty queued-thread sets as absent across filtering, list-item construction, and layout. It also skips the extra settled-thread lookup when no selected thread key exists.

Changes

Thread list behavior

Layer / File(s)	Summary
Queued-thread set handling apps/mobile/src/features/threads/threadListV2.ts	Section filtering, list-item construction, and layout classification treat an empty queued-thread set as absent. Nonempty sets use thread-key membership checks.
Selected settled-thread lookup apps/mobile/src/features/threads/threadListV2.ts	The lookup for an extra settled row runs only when a selected thread key exists. It remains limited to rows beyond the current page.

Priority: ⬇️ Low

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

Change: Refactor

Suggested reviewers: t3dotgg

Merge Risk: ⚪ Minimal · up to 4ff5c

The list optimizations preserve queued-state and selected-row behavior, so no concrete merge-blocking risk is identified.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
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 3 functions across 1 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.
Title check	✅ Passed	The title clearly and concisely describes the main change: skipping unnecessary thread-key lookups in the mobile list.
Description check	✅ Passed	The description explains the problem and change, gives detailed measurements with appropriate limitations, and reports focused tests and checks. It also notes the maintainer’s request, but does not li…

✨ Finishing Touches

📝 Generate docstrings

• Commit to this branch
• Create a new PR

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

• Autopilot · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

---

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