[tiensonqin]

Summary

Keep title sync from repeatedly scanning every node in a journal just to refresh the visible 32-row window. Preserve only the selected UUIDs across title/updated-at commits and local title edits, then hydrate their contents from the latest projection. Actual journal attributes drive eligibility; structural changes, non-title pending operations, title presence changes, and unbounded reads retain the existing full-read path.

This is the simpler follow-up to #156: no DataScript changes, extra runtime, thread, or sync-protocol changes. Apply/checkpoint/ACK/rebase/error ordering is unchanged. It shortens common sync jobs, not preemption of atomic structural work.

Validation: 1,297 core tests; native WebSocket command/recovery/cancellation checks; real local three-client sync stress (602 mobile operations, 82 CLI mutations, 20 assets, offline/crash/lost-ACK/server-restart recovery, matching final datoms, zero pending). One regression case compares cached and fresh journal reads across content and structural commits.

macOS real-core probe with 10,000 journal nodes: window refresh ~78ms → ~0.6ms, including a newer unsynced local title. This is not physical iPad profiling.

Link to Devin session: https://app.devin.ai/sessions/951fc198b3c140ad8b8cc8af5b531471
Open in Devin Desktop: https://app.devin.ai/desktop/session/951fc198b3c140ad8b8cc8af5b531471?variant=devin
Requested by: @tiensonqin

---

🟢 Devin Review found nothing to fix · View findings

[devin-ai-integration]

I'll fix CI failures and address comments from users with write access. I'll skip comments containing "(aside)".

• Disable automatic comment, CI, and merge conflict monitoring

[devin-ai-integration]

✅ Devin Review: No Issues Found

Devin Review analyzed this PR and found no bugs or issues to report.