feat(multicursor): MC HAMMER #41587

Other (squashed) commits:

fix(tui): emit ui_send output atomically with the frame

Problem:
tui_ui_send() writes directly to the TTY, bypassing the output buffer.
Sequences sent via nvim_ui_send() (e.g. kitty multiple-cursors, or
visual-dot-repeat) always arrive in a separate TTY write from the frame
they were computed for. This manifests as "tearing", or e.g. in the case
of multicursor the terminal renders text with stale cursor overlays.

Solution:
- tui_ui_send(): while a frame is being assembled (pending invalid
  regions or buffered output), buffer instead of writing directly.
  - Out-of-frame sends (tty queries, clear-on-disable) still write
    immediately.
- mcursor.lua: emit the terminal-cursor update at the end of the redraw
  cycle (`on_end`, when screen positions are final) instead of
  vim.schedule().

runtime/lua/vim/_meta/api.gen.lua

@@ -119,6 +119,12 @@ function vim.api.nvim__inspect_cell(grid, row, col) end
 --- reach, so this function can be used to force a cache clear in a test.
 function vim.api.nvim__invalidate_glyph_cache() end
 
+--- WARNING: This feature is experimental/unstable.
+---
+--- Returns true if a multicursor cascade is in-progress.
+--- @return boolean
+function vim.api.nvim__mcursor_cascading() end
+
 --- WARNING: This feature is experimental/unstable.
 ---
 --- Get the properties for namespace
@@ -1674,6 +1680,13 @@ function vim.api.nvim_list_wins() end
 --- @return any
 function vim.api.nvim_load_context(dict) end
 
+--- Adds a multicursor in the given buffer.
+---
+--- @param buf integer Buffer handle, or 0 for current buffer
+--- @param pos [integer, integer] (row, col) (1,0)-indexed cursor position (byte offset)
+--- @return integer # Total number of extra cursors.
+function vim.api.nvim_mcursor(buf, pos) end
+
 --- @deprecated
 --- @param msg string
 --- @param log_level integer