perf: count member access once and scope find highlights to painting - #3464

…iews (#3466)

## Problem

Link previews (#3219, #3225) find a message's preview URLs with
`previewLinks`. For any body with an `https://` link, that function
called `prepareMarkdown`, then scanned the prepared content again with
`scanMarkdown`, and kept no result. The renderer already prepares the
same body through its own cache.

A link-bearing row therefore parsed its Markdown two or three more times
than a row without links, on every mount. It did this again on each
remount while scrolling or switching channels.

## Change

- `markdown-preparation.ts`:
- Owns the cache of recent prepared bodies. `recentResults` and
`preparedBodies` move here from `MessageMarkdown.tsx`, unchanged.
  - Exposes it as `preparedMarkdown(content, scanned?)`.
- `prepareMarkdown` takes an optional `scanned` callback that receives
the scan it made. It stays pure and keeps no state.
- `MessageMarkdown.tsx` reads through `preparedMarkdown`, with the same
cache and budget as before.
- `preview-links.ts`:
- Uses `preparedMarkdown`. On a fresh preparation it takes the scan
through the callback instead of scanning again.
- Keeps recent preview link lists in a `recentResults` cache. The key is
the content and exclusions; the weight counts both the key and the URLs.
Cached lists are frozen.
- `preview-links.test.ts` asserts that the renderer and previews
together scan a body once. This fails on the previous code.

## Why behaviour is unchanged

- `prepareMarkdown` is pure, so a cached prepared body is identical to a
fresh one.
- The scan passed to the callback is the scan of the returned `content`,
the same input previews scanned before.
- The link list depends only on content and exclusions, which are both
in the cache key.

## Evidence

Interleaved A/B on 9 Linux and 9 macOS boxes:
- 6 rounds × 2 samples per variant on each box, so n=108 per variant per
OS.
- 4× CPU throttle.
- Values are the mean of per-box medians.

The baseline is `main` (`7bb45482`) with #3464 and #3465 applied. This
change was measured on top of that. The measured revision handed the
scan over through module state instead of a callback; the parse count is
the same. A variant with preview scanning disabled entirely is the
ceiling.

| case | metric | Linux base | Linux PR | Δ | macOS base | macOS PR | Δ
| ceiling Δ (Linux / macOS) |
|---|---|---|---|---|---|---|---|---|
| thread scroll open | script ms | 636.4 | 618.3 | −2.8% | 381.4 | 368.2
| −3.5% | −2.9% / −3.7% |
| | task ms | 997.3 | 976.1 | −2.1% | 600.6 | 588.1 | −2.1% | −1.4% /
−2.0% |
| | LoAF blocking ms | 551.7 | 534.3 | −3.1% | 271.2 | 258.4 | −4.7% |
−2.5% / −5.4% |
| | first rows ms | 937.1 | 914.1 | −2.5% | 561.5 | 546.2 | −2.7% |
−1.8% / −2.7% |
| cold markdown channel open | script ms | 495.6 | 487.8 | −1.6% | 320.6
| 316.8 | −1.2% | −2.0% / −1.7% |
| fast scroll paging (markdown) | LoAF blocking ms | 1489.8 | 1444.2 |
−3.1% | 83.7 | 83.5 | −0.3% | −2.3% / −5.3% |

- Thread scroll open improved on every box (9/9) for script and LoAF on
both OSes.
- The change recovers all the time that link-preview scanning costs in
these scenarios.
- An earlier A/B of desktop-v0.7.18 against the same baseline put thread
scroll open at +2.6% script on both OSes, which this removes.

## Checks

- `vitest run src/features/messages`: 3119 pass.
- `link-previews.spec` and `markdown-documents.spec` pass on chromium
and webkit.
- `bin/pnpm check` is clean.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Signed-off-by: Kalvin Chau <[email protected]>
Co-authored-by: Claude Opus 5.5 <[email protected]>

## Problem

#3316 added right-click message actions by wrapping every toolbar
message row in its own Base UI `ContextMenuRoot` and
`ContextMenuTrigger`. Each of those mounts costs root state, floating-ui
setup, and trigger listeners. With a full timeline that adds up to about
a tenth of list render script time, and a quarter or more of
long-animation-frame blocking while scrolling.

Bisect against v1.0.1 isolated this commit as the per-row cost. Box A/Bs
then split the cost into parts:
- A row with no menu at all recovers almost the whole regression.
- A Base UI root and trigger per row with no props costs nearly as much
as main.
- The cost is therefore the per-row Base UI menu, not the action props.

## Change

`MessageContextMenus` in `MessageActionBar.tsx` is one real Base UI
context menu per message list. It wraps the channel timeline feed and
the thread panel's rows.
- **Rows inside a host.** Each toolbar row registers its actions by
message ID, in an effect, and renders only its content.
- **Choosing the menu.** A right-click, or the ContextMenu / Shift+F10
key, over a registered row selects that row's actions. Presses outside a
registered row keep the browser's native menu, as do the existing native
targets (links, media, fields, selection).
- **Gestures.** Base UI still owns every gesture: mouseup guards,
outside-press dismissal, and switching to another row while a menu is
open.
- **Focus.** The popup renders inside the list's React tree, so the
timeline's focus bookkeeping still sees focus enter a row's menu, as it
did with a row's own menu.
- **Updates and removal.** The open menu follows its row's latest
actions and closes if the row unmounts. The host releases the row when
the menu closes.
- **Rows outside a host** keep their own menu: standalone uses, media
review, activity and huddle. Coordination rows keep their own trigger
too, and Base UI stops their event before it reaches the host.

`ThreadPanel.tsx` only wraps the rows element. Its diff is indentation;
`git diff -w` shows 3 lines.

## Evidence

Interleaved A/B on 9 Linux and 9 macOS boxes:
- 6 rounds × 2 samples per variant on each box, so n=108 per variant per
OS.
- 4× CPU throttle.
- Values are the mean of per-box medians.

The baseline is `main` (`7c786bb4`) with #3464 applied (`90580d91`).
This change was measured stacked on top of it (`3caa47b0`). The final
commit differs from the measured one only in menu open, close and focus
handling, which these scenarios do not exercise.

| case | metric | Linux base | Linux PR | Δ | macOS base | macOS PR | Δ
|
|---|---|---|---|---|---|---|---|
| fast scroll paging (markdown) | script ms | 6148.8 | 5600.0 | −8.9% |
3859.6 | 3601.6 | −6.7% |
| | task ms | 11384.7 | 10831.0 | −4.9% | 7206.1 | 6907.9 | −4.1% |
| | LoAF blocking ms | 1849.6 | 1475.8 | −20.2% | 138.3 | 82.4 | −40.4%
|
| channel open with hover | script ms | 273.1 | 246.2 | −9.9% | 162.9 |
150.2 | −7.8% |
| | task ms | 465.7 | 438.7 | −5.8% | 276.0 | 262.2 | −5.0% |
| | LoAF blocking ms | 260.3 | 242.4 | −6.9% | 126.0 | 118.9 | −5.6% |
| cold channel open | script ms | 292.7 | 267.6 | −8.5% | 184.8 | 172.2
| −6.8% |
| | task ms | 515.5 | 492.4 | −4.5% | 332.9 | 320.4 | −3.7% |
| | LoAF blocking ms | 245.1 | 230.1 | −6.1% | 106.9 | 98.6 | −7.7% |
| rich thread open | script ms | 2417.1 | 2275.4 | −5.9% | 1138.8 |
1102.8 | −3.2% |
| | task ms | 5020.0 | 4842.7 | −3.5% | 2385.8 | 2356.3 | −1.2% |
| | LoAF blocking ms | 2335.3 | 2175.1 | −6.9% | 895.8 | 852.3 | −4.8% |
| warm channel switch | script ms | 133.1 | 129.3 | −2.9% | 91.0 | 89.8
| −1.3% |
| | LoAF blocking ms | 116.2 | 112.3 | −3.4% | 49.0 | 48.3 | −1.3% |

All 9 boxes improved on both OSes for every row except warm channel
switch, which is small and mixed (Linux script 9/9, the rest 6–7/9).
Style time is unchanged (±1.3%).

## Checks

- `vitest run src/features/messages src/features/conversation`: 3286
pass. New unit tests cover:
  - which row's actions open
  - the native menu outside rows
  - closing when the row is removed
  - keyboard open
  - focus staying in the list
- The removal and focus tests were each confirmed to fail without their
fix.
- Browser specs pass on chromium and webkit: `message-context-menu`,
`menu-dismiss`, `message-actions`, `message-actions-floating`,
`buzz-links`, `external-links`, `thread-composer-focus`,
`find-on-screen`.
- `tsc --noEmit` and `biome check` are clean.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Signed-off-by: Kalvin Chau <[email protected]>
Co-authored-by: Claude Opus 5.5 <[email protected]>

* origin/main: (153 commits)
  feat(agents): preview where interrupted agents were working before resume (#3494)
  perf(messages): close the editor and delete dialog on submit (#3493)
  feat(messages): quote selected message text into the composer (#2392)
  perf(toast): scope docked clearance to the toast host (#3489)
  fix(unread): undo exact captured ordinary reads conditionally (#3473)
  fix(solo): keep navigation responsive during chat deletion (#3488)
  fix(channels): keep custom thread tab names until the tab closes (#3458)
  perf(relay): verify cached history in the native host (#3426)
  fix(settings): refine claude and typesafe key setup dialogs (#3391)
  Update bundled help for Bestie and agent settings (#3194)
  feat(profiles): share and show a local time zone (#3332)
  Revert "feat(link-previews): unfurl web links below messages by default" (#3480)
  Show disabled plugin chats in Solo (#3456)
  perf(messages): host one context menu per message list (#3465)
  test(claude): keep the quit-release test copy above the minimum version (#3477)
  fix(profiles): restore verified names from the device cache (#3382)
  fix(mentions): show addressed agents first in Solo audience picker (#2856)
  perf: count member access once and scope find highlights to painting (#3464)
  perf(messages): prepare a message once for its renderer and link previews (#3466)
  fix(composer): speed up long pastes and paste office text as text (#3441)
  ...

# Conflicts:
#	docs/plugin-authoring-guide.md

…when-ready

* origin/main: (21 commits)
  fix(agents): report a conversation wake that fails before its turn starts (#2192)
  fix(usage): read Claude transcripts over 50 MB a page at a time (#3498)
  feat(agents): preview where interrupted agents were working before resume (#3494)
  perf(messages): close the editor and delete dialog on submit (#3493)
  feat(messages): quote selected message text into the composer (#2392)
  perf(toast): scope docked clearance to the toast host (#3489)
  fix(unread): undo exact captured ordinary reads conditionally (#3473)
  fix(solo): keep navigation responsive during chat deletion (#3488)
  fix(channels): keep custom thread tab names until the tab closes (#3458)
  perf(relay): verify cached history in the native host (#3426)
  fix(settings): refine claude and typesafe key setup dialogs (#3391)
  Update bundled help for Bestie and agent settings (#3194)
  feat(profiles): share and show a local time zone (#3332)
  Revert "feat(link-previews): unfurl web links below messages by default" (#3480)
  Show disabled plugin chats in Solo (#3456)
  perf(messages): host one context menu per message list (#3465)
  test(claude): keep the quit-release test copy above the minimum version (#3477)
  fix(profiles): restore verified names from the device cache (#3382)
  fix(mentions): show addressed agents first in Solo audience picker (#2856)
  perf: count member access once and scope find highlights to painting (#3464)
  ...

…n-on-github

* origin/main: (21 commits)
  fix(agents): report a conversation wake that fails before its turn starts (#2192)
  fix(usage): read Claude transcripts over 50 MB a page at a time (#3498)
  feat(agents): preview where interrupted agents were working before resume (#3494)
  perf(messages): close the editor and delete dialog on submit (#3493)
  feat(messages): quote selected message text into the composer (#2392)
  perf(toast): scope docked clearance to the toast host (#3489)
  fix(unread): undo exact captured ordinary reads conditionally (#3473)
  fix(solo): keep navigation responsive during chat deletion (#3488)
  fix(channels): keep custom thread tab names until the tab closes (#3458)
  perf(relay): verify cached history in the native host (#3426)
  fix(settings): refine claude and typesafe key setup dialogs (#3391)
  Update bundled help for Bestie and agent settings (#3194)
  feat(profiles): share and show a local time zone (#3332)
  Revert "feat(link-previews): unfurl web links below messages by default" (#3480)
  Show disabled plugin chats in Solo (#3456)
  perf(messages): host one context menu per message list (#3465)
  test(claude): keep the quit-release test copy above the minimum version (#3477)
  fix(profiles): restore verified names from the device cache (#3382)
  fix(mentions): show addressed agents first in Solo audience picker (#2856)
  perf: count member access once and scope find highlights to painting (#3464)
  ...