fix(web): keep sidebar drag dividers clear and gestures smooth

[juliusmarminge]

Sidebar drag dividers overlapped thread content, could remain visible after interrupted gestures, and disappeared behind cards dragged into empty Pins.

Dividers and empty-section targets now reserve clearance only during a drag. Empty Active previously left a hidden 36px target behind when the last active thread was pinned; the target now stays zero-height at rest. Rows retain smooth sorting transitions. Crossing the visible divider row immediately switches between Pins and Active. The target follows pointer movement, so opening a slot cannot toggle a stationary gesture back. The lifted card stops below the Pins label at the top edge. A single gesture cleanup path handles release, Escape, window blur, hidden tabs, missed releases, and list unmount, without turning a cancelled drag into a thread click.

Verified with 237 focused sidebar tests, web typecheck, and browser checks for empty Pins, full and compact rows, successful drops, cancellation, and subsequent drags. Targeted lint and React Doctor report existing sidebar warnings. The change applies to the sidebar shared by web and desktop; mobile and wire contracts are unchanged.

Matching 1280 × 720 dark-mode captures use identical disposable synthetic threads. The static spacing comparison is main at 52b2bf77a94 versus d0ec56e58a7. The divider-crossing recording is from 0422f94ed06. The empty-target spacing comparison below covers the follow-up at 4aa3a090707.

Before: dragging within Settled	After: dragging within Settled

Before: after window blur	After: after window blur

Divider-crossing recording: cross the visible divider into Pins, cross it back into Active, and return to Pins to drop. A one-pixel movement onto the divider changes the target; no thread-center threshold is involved. Empty Pins and compact settled rows were checked separately, including top-edge clamping and release cleanup.

https://gh-file-drop-api-prod-mi5fy3sowv63ufte.pinglabs.workers.dev/f/ee332894b7e464c2/cross-divider.mp4

Empty Active regression: with identical pinned threads, the old target retained 36px even after all labels and transforms cleared. Three repeated drag/drop cycles reproduced it; the same three cycles with this fix ended at zero target height. No separate stuck-transform case was reproduced.

Before: hidden target retains space	After: empty target collapses

Macroscope skipped correctness review because the workspace monthly spending limit was reached; its approvability summary reports that limitation. The full diff was reviewed locally.

Authored with GPT-6 in Codex.

[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	13.6 KiB	13.5 KiB	−70 B (−0.5%)	15.1 KiB	✅
Codex	Thread snapshot wire	7.0 KiB	7.0 KiB	−2 B (−0.0%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.6 KiB	6.5 KiB	−68 B (−1.0%)	7.8 KiB	✅
Codex	Live turn WebSocket decoded	57.1 KiB	57.0 KiB	−88 B (−0.2%)	66.4 KiB	✅
Codex	Live turn messages	10	8	−2 (−20.0%)	21	✅
Claude	Total thread wire	13.6 KiB	13.6 KiB	+8 B (+0.1%)	15.1 KiB	✅
Claude	Thread snapshot wire	7.0 KiB	7.1 KiB	+1 B (+0.0%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.5 KiB	6.5 KiB	+7 B (+0.1%)	7.8 KiB	✅
Claude	Live turn WebSocket decoded	57.8 KiB	57.8 KiB	0 B (0.0%)	66.4 KiB	✅
Claude	Live turn messages	8	8	0 (0.0%)	21	✅

Baseline: 52b2bf7 · PR result: 4aa3a09 · 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.

[macroscopeapp]

Approvability

Verdict: Not approved

Macroscope's review found this PR not approvable — The change substantially rewires the shared sidebar’s drag sensor, cancellation lifecycle, collision targeting, layout projection, and visual spacing. Although focused tests cover many scenarios, the breadth and stateful nature of the production interaction changes warrant 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.}