Bind the worktree head's keys through DISCLOSE, and name the tab strip's own press

[leaf-agent]

Problem

Two leftovers from 335bf9a "Delegate page chrome to browser standards", which moved control activation off a page-wide CONTROL scope and into rows declared where each control is. Both are on ci's red list for the run this was dispatched for or reachable from the same change, and both still reproduce at 1ca4f40.

lf-worktree's head names fewer keys than work it. The commit gave the head a row binding PRESS — Enter and Space — over a button wearing aria-expanded, which is ARIA's disclosure pattern. skills/leaf/CLAUDE.md states what that costs: "every row over a disclosure reads [DISCLOSE] — this scope's, and a widget's own row re-wording the same press. Two rows naming different sets is not two promises but one: lineRows prints the nearer row and drops the other whole, so a widget naming one key fewer takes the rest off the line". The widget row is nearer, so the arrow that opens the evidence went on working and stopped being named — on the key line and in aria-keyshortcuts, which is written once when the scope is declared and so froze at Enter Space. Nothing tested it; the identical bug in lf-options is what #132 is for, and that PR does not reach this package.

test_a_widget_that_renames_its_role_keeps_the_press_offer_gave_it still asks for the deleted scope's word. It asserts the key line reads press it over a focused tab. press it was CONTROL's line, and 335bf9a removed the scope in the same commit that gave lf-tabs its own tab.activate row reading open the tab. The commit updated several tests in that file and dropped test_the_reference_names_the_space_that_works_a_control outright; this assertion was missed, so the test fails before reaching the two facts it exists to hold.

Solution

lf-worktree's head takes its keys from DISCLOSE(head), the way lf-diff and lf-draft already do, and keeps its own run. DISCLOSE hands over only the arrow that changes the state, so the press stays a direction rather than a second toggle; it also answers for where the head stands, and the runtime's disclosure scope stops at the chrome. The head is a span from selectableOffer, so a copy frozen into thread markup has no platform pair underneath it the way a details > summary does — the row's own press is the only thing there, which is the shape #132 keeps for the sibling widget. Its line reads the way the tree is standing, and aria-expanded is written at birth so the head is never briefly a control the runtime cannot place — without that write DISCLOSE hands back both arrows, and that is the value aria-keyshortcuts is written with when the scope is declared. The repaint that keeps the line and aria-keyshortcuts together is the document's disclosure watch, which #132 moved from paintHere() to paintKeys() while this branch was open; the widget owes none of its own, which is what skills/leaf/CLAUDE.md now says in as many words.

The tabs test asserts open the tab — the strip's own word for the same press — and its docstring now describes where that press is declared rather than the scope that used to supply it. Everything after the assertion is unchanged: Enter still has to re-select the focused tab, and Space still has to be consumed rather than scrolling the page out from under it.

New coverage in test_render_projection.py for the worktree head, since nothing exercised it. It holds the row rather than the repaint: on the page, both surfaces naming the row's keys — the line and aria-keyshortcuts — plus a second → on an open tree, so a scope answering nothing at all could not pass it. In a message: the same head in the threads panel, where aria-keyshortcuts reads Enter Space, the arrow moves nothing, and Enter and Space fold it. The repaint behind both surfaces is held by test_a_widgets_native_control_names_the_press_the_platform_makes, which #132 added.

Testing

uv run pytest tests at c292244 — 723 passed, 6 skipped. After merging main (which brought #132 and #143) and taking the repaint out at 890e7ff, test_render_projection.py, test_render_navigation.py, and test_render_anchors.py — 193 passed, 4 failed, all four the version-menu group #137 owns and all four failing identically on origin/main in a clean worktree. With --run-nightly, test_render_projection.py and test_render_navigation.py — 137 passed, 1 failed, that one being test_the_arrows_say_which_way_the_section_under_the_reader_goes, which is #132's settled ask and untouched here; it fails the same way at the merge base 1ca4f40.

Each half of the new test was run against the code it holds and fails there: the page half at assert 'Enter Space' == 'Enter Space ArrowRight' on main's lf-worktree.js, and the thread-markup half at aria-expanded staying "false" under Enter with the row's run removed.

What this does not fix in that run

The run reported 34 failures. Most of the rest already have a PR open: the -15 root scrollbar-gutter family is #121 (assert -15 == 0, the drawn edge landing a bar's width off the hand, the covering panel's blank strip), the version-menu group is #137, the settled ask's disclosure keys were #132, now merged, the margin and float placements are #138, the gallery stills are #134, and the sidebar's stated window is #136.

No guard stops the third widget from declaring a fixed key pair over a disclosure. DISCLOSE exists so the two surfaces cannot come apart, but a widget has to remember to call it, and two of the four widgets over a disclosure did not. A corpus-wide check — every focusable disclosure names the keys DISCLOSE gives it — would catch the next one, and is left out of this PR because it belongs beside the corpus checks rather than in a fix for two widgets.

---

Automated fix for failed run

[leaf-agent]

The worktree head is a span[role="button"] from selectableOffer, not a native disclosure, so dropping run takes its press away wherever the runtime's DISCLOSURE scope does not reach. That scope is standingOn(...), whose at reads Boolean(el?.matches?.(sel)) && !inChrome(el), and nothing has bound a press over [data-lf-offer] since 335bf9a retired CONTROL — so on a head frozen into thread markup the row names ⏎ / space and nothing runs them.

lf-diff and lf-draft survive the same shape because their disclosure is details > summary and the platform answers the pair there. test_the_arrows_say_which_way_the_section_under_the_reader_goes says so about the staged diff in as many words: "The platform's pair still works it, so what differs is the offer rather than the capability." A span has no platform half to fall back on.

Measured on this branch and on the base, appending <lf-roster><lf-agent><lf-worktree source="atlas-worktrees"> as a comment's markup and focusing the head in the threads panel:

	aria-keyshortcuts	Enter	Space
base 1ca4f40	Enter Space	opens	closes
15d18b7	Enter Space	nothing	nothing

The premise for dropping the binding — that a run over the arrows would make them a second toggle rather than a direction — does not hold. DISCLOSE only ever offers the arrow that changes the state, which is what the runtime's own row already says: "One toggle answers all three: the arrow bound is the one that changes this disclosure, so a press cannot mean anything else." Its run is focused().click(), a toggle. #132 reaches the same conclusion for the sibling widget and keeps run: () => this.#row.click() beside its keys: () => DISCLOSE(this.#row).

With run: () => head.click() restored, test_worktree_evidence_names_the_arrow_that_stands_on_it still passes whole — the second → on an open tree included, since the widget row binds only the arrow DISCLOSE named — and the head answers Enter and Space in the panel again. Nothing in the suite covers a worktree head in chrome, which is why the new test could not see this; a case beside the staged-diff one in test_the_arrows_say_which_way_the_section_under_the_reader_goes is where it would live.

[leaf-agent]

The run restoration in c292244 holds up — I removed it and test_worktree_evidence_names_the_arrow_that_stands_on_it fails at the frozen head's Enter, so that half of the widget row is genuinely covered. Resolving that thread.

Two findings on what the same commit does not cover, and where the repaint it adds belongs. Both come from deleting the line and re-running, on this branch's merged tree.

The paintKeys() call and the birth aria-expanded write are each unproven by the new test: delete either one and the test stays green (3/3 for the paintKeys() deletion). The docstring claims the opposite in as many words, and takes a deliberate precaution — reading aria-keyshortcuts once rather than retrying — against the two-second heartbeat. The masking repaint is not the heartbeat: it is paintAnchors → paintThreadQuotes → paintKeys, which runs on the same gesture, so aria-keyshortcuts is already correct on the read taken immediately after the press, with the widget's own writes removed.

And the repaint sits a layer below where skills/leaf/CLAUDE.md puts it. Open #132 moves disclosureWatch from paintHere() to paintKeys() so the document's watch — which already hears this exact aria-expanded write — keeps both surfaces together, and adds the sentence "a widget declaring a disclosure row owes no repaint of its own". Landing both leaves lf-worktree carrying a local repaint that the doc arriving with #132 says it does not owe, and a second full scope revalidation per toggle behind the watch's.

Separately, and worth saying because it raises this PR's urgency rather than lowering it: the test_a_widget_that_renames_its_role_keeps_the_press_offer_gave_it change is a real red-main fix, not test polish. On the merged tree the base assertion fails — the key line reads ⏎ / space open the tab, which is lf-tabs' own tab.activate row from 335bf9a, and nothing in the tree says "press it" any more. main still carries the old assertion.

How each was measured

All on the PR's merged tree (c292244 + base), uv run pytest against the sandbox's own browser.

• run removed from the widget row → test_worktree_evidence_names_the_arrow_that_stands_on_it fails.
• paintKeys() removed → passes, 3 runs of 3.
• Birth head.setAttribute("aria-expanded", …) removed → passes.
• Probe with paintKeys instrumented to record its stack: one call per gesture, paintAnchors → paintThreadQuotes → paintKeys, and aria-keyshortcuts already reads Enter Space ArrowLeft on the read taken immediately after the press (and after a pointer click) with the widget's paintKeys() gone.
• Base assertion restored in test_a_widget_that_renames_its_role_keeps_the_press_offer_gave_it → fails; actual key line is ⏎ / spaceopen the tabesclet go← / →walk the tabs….