W14: pin K09/K11/K39 landed W5 special-form fixes (test-only)

Three dc7aa709 fixes shipped without pinning tests:
- K09: R7RS longhand (unquote-splicing X) now splices (was silent zero-splice)
- K11: guard re-raise sentinel gensym'd — a user value shaped like
  (list '__guard-reraise__ X) is data, not a forged re-raise
- K39: (do ((fn (x) x) 5) 99) -> 99, not a misparsed Scheme do-loop

Add suites gate-K09-longhand-unquote-splicing, gate-K11-guard-reraise-forgeable,
gate-K39-do-iife-head to spec/tests/test-gate-pins.sx with exact reprs from
plans/sx-review/core.md. 261 passed / 0 failed under OCaml run_tests.

Test-only: no semantics edits, no push.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

spec/tests/test-gate-pins.sx

@@ -9,6 +9,7 @@
 ;;
 ;; TEST-ONLY: no semantics edits. If a pin fails, the fix regressed — do NOT
 ;; relax the assertion; investigate the evaluator/primitive change.
+;; NB: assert= uses `=` (not `equal?`); compare lists with `=`.
 ;; ==========================================================================
 
 ;; --------------------------------------------------------------------------
@@ -56,3 +57,65 @@
   (deftest
     "contains? still works on string substrings"
     (assert (contains? "hello" "ell"))))
+
+;; --------------------------------------------------------------------------
+;; K09 [W5, high] R7RS longhand (unquote-splicing X) silently no-spliced —
+;; only shorthand ,@/`splice-unquote` was recognized, so the longhand
+;; serialized literally (zero-splice). Fixed: aliased to splice-unquote.
+;; Repro (core.md).
+;; --------------------------------------------------------------------------
+(defsuite
+  "gate-K09-longhand-unquote-splicing"
+  (deftest
+    "longhand unquote-splicing splices a list"
+    (assert=
+      (quasiquote
+        (1
+          (unquote-splicing (list 2 3))
+          4))
+      (list 1 2 3 4)))
+  (deftest
+    "longhand unquote-splicing of an empty list contributes nothing"
+    (assert=
+      (quasiquote (0 (unquote-splicing (list)) 9))
+      (list 0 9)))
+  (deftest
+    "shorthand splice-unquote still works"
+    (assert=
+      (quasiquote (a (splice-unquote (list 2 3)) z))
+      (list (quote a) 2 3 (quote z)))))
+
+;; --------------------------------------------------------------------------
+;; K11 [W5, high] guard re-raise sentinel was a plain forgeable symbol — a
+;; body/clause legitimately returning (list '__guard-reraise__ X) was
+;; misread as a re-raise of X. Fixed: sentinel gensym'd per execution, so a
+;; user value with that head is returned as data. Repro (core.md).
+;; --------------------------------------------------------------------------
+(defsuite
+  "gate-K11-guard-reraise-forgeable"
+  (deftest
+    "body value shaped like the sentinel is returned as data"
+    (assert=
+      (guard (e (true "caught")) (list (quote __guard-reraise__) "hi"))
+      (list (quote __guard-reraise__) "hi")))
+  (deftest
+    "clause returning the forged sentinel is not re-raised"
+    (assert=
+      (guard
+        (e (true (list (quote __guard-reraise__) "forged")))
+        (error "boom"))
+      (list (quote __guard-reraise__) "forged"))))
+
+;; --------------------------------------------------------------------------
+;; K39 [W5, med] `do` misparsed a first form whose head is a list (an IIFE)
+;; as a Scheme do-loop binding spec. Repro (core.md): (do ((fn (x) x) 5) 99)
+;; threw "first: expected list, got 5"; expected 99. Fixed: `do` is begin.
+;; --------------------------------------------------------------------------
+(defsuite
+  "gate-K39-do-iife-head"
+  (deftest
+    "do with an IIFE first form returns the last form (not a do-loop)"
+    (assert= (do ((fn (x) x) 5) 99) 99))
+  (deftest
+    "do with a single IIFE form returns its value"
+    (assert= (do ((fn () 42))) 42)))