fix(chat): stop pushing follow-up messages to the top

[t3dotgg]

Follow-up messages jumped to the top of a thread and left most of the screen empty.

Only the first user message now gets top anchoring and reserved space. Follow-ups stay at the real bottom on web, desktop, and mobile, including plan follow-ups.

Tests: 48 focused tests, web/mobile/shared type checks, targeted lint.

Model: GPT-5.6 Sol
Harness: Codex

---

Note

Medium Risk
Changes shared chat-list anchoring and send-time scroll on web and mobile, so a regression could mis-scroll threads or leave empty reserved space. Logic is covered by focused tests but still touches a high-traffic UX path.

Overview
Stops follow-up sends from jumping to the top of the thread and leaving a large empty gap. Only the first user message still gets reserved end-space; later sends stay at the live bottom on web and mobile.

resolveChatListAnchoredEndSpace now walks from the start and anchors only the first eligible user row. Web ChatView sets that first-message anchor when a thread has no turn and no user messages yet; otherwise it scrollToEnd(). Mobile splits submit vs. layout: submittedMessageId drives post-send scroll/follow reset, while resolveThreadFeedSubmissionAnchor keeps the first-message layout pin through queued outbox sends.

Also treats a connecting phase as not yet acknowledging a local dispatch so follow-ups stay active while the provider session starts.

^{Reviewed by Cursor Bugbot for commit 6d72b36. Bugbot is set up for automated code reviews on this repo. Configure here.}

Note

Anchor only the first user message and scroll follow-ups to end

• Introduces resolveThreadFeedSubmissionAnchor to decide at submission time whether a message gets an anchor: only the first user message in a not-yet-started turn qualifies; follow-up sends in an active turn get no anchor and rely on end-follow scrolling.
• Updates both mobile (ThreadDetailScreen, ThreadFeed) and web (ChatViewContent, MessagesTimeline) send paths to use the new rule, tracking submittedMessageId instead of anchorMessageId for scroll-to-end behavior.
• Reworks resolveChatListAnchoredEndSpace to scan forward and reserve end space only for the first eligible user-role message; later eligible rows are ignored.
• Fixes hasServerAcknowledgedLocalDispatch to return false during the 'connecting' phase, preventing premature acknowledgement of follow-ups while a provider session is starting.
• Risk: resolveChatListAnchoredEndSpace now returns undefined when the anchor id does not match the first eligible user message; any caller expecting reverse-scan semantics or anchoring of later rows will no longer reserve end space.

^{Macroscope summarized 6d72b36.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: f04b7be7-93de-418c-bc46-739448548e16

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

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[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	13.4 KiB	−26 B (−0.2%)	15.1 KiB	✅
Codex	Thread snapshot wire	6.9 KiB	6.9 KiB	−3 B (−0.0%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.6 KiB	6.5 KiB	−23 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	13.4 KiB	13.4 KiB	−1 B (−0.0%)	15.1 KiB	✅
Claude	Thread snapshot wire	6.9 KiB	6.9 KiB	−12 B (−0.2%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.5 KiB	6.6 KiB	+11 B (+0.2%)	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: ce91284 · PR result: 6d72b36 · 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.

[macroscopeapp]

Approvability

Verdict: Skipped

Macroscope did not run approvability analysis for this PR. Macroscope could not determine whether this PR modifies its approvability configuration, so the PR was not approved automatically. A PR that may change the rules that govern approval is never approved automatically.

[cursor]

Cursor Bugbot has reviewed your changes using high effort and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit fba3c7c. Configure here.}