fix(web): stop collapsing the composer when it loses focus

[t3dotgg]

Clicking a message, copying a code block, or selecting text for a citation collapsed the composer. Blur is a side effect of doing anything else on the page, not a request for more reading room. It shipped as a default in #7855 with no prior discussion, drew complaints within hours, and #9469 added a setting to turn it off instead of fixing the default.

The desktop composer now rests only when you scroll the conversation. Focus and clicks elsewhere leave it alone. The blur trigger is removed rather than defaulted off: the composerCollapseOnBlur setting, the multi-select in Settings, the selection-hold helper, and the document-level focus and pointer listeners that existed only to detect blur are gone. The Collapse composer on scroll switch stays. A stored composerCollapseOnBlur value is dropped on decode.

Mobile keeps its own focus-driven collapse, which exists for the keyboard and is unrelated.

Verification: vp test run on the contracts, desktop settings, composer layout, focus state, and timeline tests (185 passing), tsgo --noEmit in contracts, web, and desktop, and lint and format on the touched files.

Created with Claude Fable 5.1 in Claude Code.

🤖 Generated with Claude Code

[macroscopeapp]

Approvability

Verdict: Not approved

Macroscope's review found this PR not approvable — This changes the default desktop composer behavior by removing blur-triggered collapsing and retiring its persisted setting, while retaining scroll-triggered collapsing. Because the change affects existing user-visible defaults and settings behavior, it warrants human review.

Not approved because:

• Monthly spending limit reached (workspace setting). Approvability relies on correctness review in order to determine eligibility

^{Review your spending limits in Billing settings. You can add or adjust custom eligibility rules. Learn more.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ The exact PR base did not have a successful artifact. Baseline uses the latest successful main measurement shown below.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	13.3 KiB	13.6 KiB	+282 B (+2.1%)	15.1 KiB	✅
Codex	Thread snapshot wire	6.9 KiB	7.0 KiB	+155 B (+2.2%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.4 KiB	6.6 KiB	+127 B (+1.9%)	7.8 KiB	✅
Codex	Live turn WebSocket decoded	55.6 KiB	57.1 KiB	+1.4 KiB (+2.6%)	66.4 KiB	✅
Codex	Live turn messages	11	10	−1 (−9.1%)	21	✅
Claude	Total thread wire	13.4 KiB	13.6 KiB	+241 B (+1.8%)	15.1 KiB	✅
Claude	Thread snapshot wire	6.9 KiB	7.0 KiB	+144 B (+2.0%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.5 KiB	6.5 KiB	+97 B (+1.5%)	7.8 KiB	✅
Claude	Live turn WebSocket decoded	56.4 KiB	57.9 KiB	+1.4 KiB (+2.6%)	66.4 KiB	✅
Claude	Live turn messages	11	10	−1 (−9.1%)	21	✅

Baseline: 42a8fd5 · PR result: bf64483 · 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: 113.9 KiB
• Claude decoded thread snapshot: 114.6 KiB

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