Merge origin/main (v1.64.0.0) into garrytan/time-attack-fork-review

Both waves fixed several of the same bugs; resolutions keep whichever
shape this branch's tests pin (#2018 jq bind, #1798 set-- pattern,
stop-ack, lock errors, polyfill windowsHide) and take main's richer
codex Step 2A (it absorbed the same mktemp fix). True unions: memory-
ingest keeps main's capability-probed --include-gitignored inside our
GIT_CEILING defense; setup wraps main's Playwright platform override in
our stale-healing install lock; package.json takes main's diff@^9 and
the combined test glob (design/test + ios-qa/daemon/test, 30s timeout).
Generated SKILL.md files regenerated from resolved templates, never
hand-picked. Ship goldens refreshed; parity/carve budgets re-measured
for the summed preamble growth of both waves (itemized per entry).

test/helpers/parity-harness.ts

@@ -214,7 +214,9 @@ const MONOLITH_INVARIANTS: ParityInvariant[] = [
     // codexPreflight() block (install + auth tri-state + CODEX_MODE branch prose),
     // landing ~6.3% over the v1.53.0.0 baseline. Intentional: it adds proper
     // not-installed vs not-authed handling, not slop.
-    maxSizeRatio: 1.08,
+    // v1.64+v1.65 merge: both waves grew the shared preamble (evidence
+    // directive + telemetry failure flags); measured 1.094.
+    maxSizeRatio: 1.10,
     minBytes: 70_000,
   },
   {
@@ -225,8 +227,9 @@ const MONOLITH_INVARIANTS: ParityInvariant[] = [
     // P3 loop tip) is added to every skill's shared preamble — intentional, ~1KB.
     // Fork port wave 2: the shared coverage-audit detection block gained the
     // multi-ecosystem markers (Django/JVM, script/target/test-file census —
-    // e3259078 port); measured 1.111x. Tight headroom only.
-    maxSizeRatio: 1.12,
+    // e3259078 port); measured 1.111x. v1.64+v1.65 merge sums both waves'
+    // preamble growth; measured 1.125.
+    maxSizeRatio: 1.13,
     minBytes: 50_000,
   },
   {