Add Utilities, I18N, Debug, Lava, and GetModOptionsCopy to the
System tables in luaui/system.lua and luarules/system.lua so that
widgets and gadgets can access them after detach-bar-modules moves
them off the Spring table.

Also create .emmyrc.json (the EmmyLua analyzer config) with the
detached modules in the globals list, plus type stubs for LSP/CLI
support. The .emmyrc.json content matches what
vscode-recommended-extensions ships, with 5 extra globals
(Utilities/Debug/Lava/GetModOptionsCopy/I18N) that only become
real top-level identifiers after detach-bar-modules runs. When
vscode-recommended-extensions has already merged, -Xtheirs in the
cherry-pick keeps this version (the superset).

# Conflicts:
#	.emmyrc.json

Restructures the 20 files under luaui/Tests/, luaui/TestsExamples/,
plus the headless-only common/testing/infologtest.lua, from bare-
global hook declarations to a return-table shape. Updates the
dbg_test_runner widget to read test hooks from the returned table.

Motivation: the pre-existing shape required the test files to run
under setfenv(chunk, testEnvironment) and define `function test()`,
`function setup()`, etc. as bare module-level globals that setfenv
redirected into the environment. That works at runtime but emmylua
can't model the sandboxing — it sees 20+ files declaring project-
wide globals like `test`, `setup`, `skip`, `cleanup`. To keep
emmylua happy, .emmyrc.json had to blacklist both test directories
under workspace.ignoreDir — a kludge on clearly-ours code. Lives on
its own leaf so the convention change can be discussed in isolation.

Minimal shape change per file — just prepend `local` to each top-
level `function` declaration, and append a final `return { ... }`
block listing whichever lifecycle hooks (skip/setup/test/cleanup)
that file actually defines. Original indentation and formatting
preserved (no stylua reformatting noise — the fmt transform runs
after this one in the mig pipeline).

Runner patch — luaui/Widgets/dbg_test_runner.lua, loadTestFromFile:
  - capture the return value of pcall(chunk)
  - require it to be a table
  - merge its keys into testEnvironment so runTestInternal still
    reads bare `skip`/`setup`/`test`/`cleanup` under setfenv

Vendored LuaCATS annotations for busted/luassert to provide
IntelliSense for the unit-test surface. Lives on its own leaf so the
discussion around 'vendoring LuaCATS types' can happen in isolation —
prior pushback on the same direction in an earlier unit-testing PR
makes this the right place to litigate it rather than burying it in a
broader env commit.

Why vendored instead of declared as a Lux dep: Lux does not yet
support pulling LuaCATS annotations from library deps, and quick
attempts to wire this up in Lux failed. Upstream tracking issue:
lumen-oss/lux#953 — once that lands, these
directories should be deleted in favor of declaring busted as a
normal Lux dev-dep.

Sources (pinned SHAs):
  - types/busted/   https://github.com/LuaCATS/busted
                    @ 5ed85d0e016a5eb5eca097aa52905eedf1b180f1
  - types/luassert/ https://github.com/LuaCATS/luassert
                    @ d3528bb679302cbfdedefabb37064515ab95f7b9

See types/busted/provenance.md and types/luassert/provenance.md for
per-directory upstream refs + license status.

…ixes

Human-curated environment that, together with the LLM type-triage pass
(fmt-llm), drives emmylua_check to zero on the migrated tree. Per-change
rationale lives in PR beyond-all-reason#7447 review comments.

- .emmyrc.json globals + diagnostics; types/* stubs; busted mock; CI gate
- forward-decl / assertEqual declarations; reverted orphaned kikito loader
- rationale-comment strip; types/IntegrationTests rename
- deterministic pins for type-triage leftovers the LLM mishandles on big
  files (multi_attack opts, HighlightUnit forward-decl, ripairs suppress,
  gui_pip gameFrame use-before-declare)
- json.lua forward-decl tidy (relocate null, drop dead decode_scan*)

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

IsDevModeCached (upstream beyond-all-reason#6918) references `utilities` inside its own
table constructor, where the local is not yet in scope — Lua resolves
those reads as GLOBALS, so the first real call would index nil. Dormant
today only because nothing calls it. Forward-declare the local so the
closure captures it.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

0.22.0's analyzer resolves a local inside its own table constructor, so
constructor-self-reference globals — a real dormant-crash class (see the
springFunctions.lua fix) — passed CI silently. 0.24.0 catches them; the
whole workspace surfaces exactly the four occurrences of that one bug,
fixed in the previous commit, so the stricter gate lands green. Also
tracks the upstream release asset rename (arm64 -> aarch64).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

gui_flowui.lua: Draw.Button uses `opaque` — Draw.Element's 17th
parameter, which Button never had; the global read was always nil.
Pinned false (behavior-identical) until upstream decides whether Button
should expose an opaque mode.

snd_notifications.lua: `customNotifications` is persisted by
GetConfigData but never declared or assigned anywhere. Declared nil so
the round-trip is explicit.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Generated by parallel claude-sonnet-4-6 workers dispatched by
scripts/codemod/llm-type-triage.sh, applying fixes per SKILL.md categories.
Single pass, no iteration — categories that don't shrink the count
are a signal that SKILL.md needs a new rule.

…swap)

…CI workflow)