engine: the verse facts are the whole scroll, derived from L1

torah_text.pl held 32 hand-written verses from Phase 1 — "the reasoning
layer's view of the specification". With the cycle and the Sanhedrin
now consulting the engine on every question (2026-10-19), that view
showed: any root it knew lit the same handful, and bereishit 1:1 and
devarim 6:4 were read in the same eleven verses.

- scripts/build_engine_verses.py writes
  talmudic_engine/src/specification/l1_verses.pl from l1/torah: every
  verse, each word's shoresh as the engine names it (Hebrew → letter
  names, finals folded), the consonantal text as gloss, and the SHA-256
  of the L1 files read in the header. `--check` exits 1 when the facts
  lag the scroll; build_and_push.sh runs it beside the Doorpost check.
- torah_text.pl includes the generated file; the hand facts are gone.
  L1's roots are authoritative where the hand facts differed (Israel
  שרה, the Name הוה).
- interpret.pl: over 5,853 verses a common root lights hundreds, so the
  reading ranks verses by how many of the case's roots each carries and
  keeps the top 12 in canonical order; the coupling trace names only the
  verses the reading cites.

Local, over the full scroll: bereishit 1:1 → itself, Devarim 4:32,
Bereishit 2:4 …; devarim 6:4 → itself, Shemot 3:18, Devarim 4:1 …;
shabbat roots → Bereishit 2:3, Shemot 20:8, 20:11, 31:13. interpret
15 ms, challenge 10 ms (now finds a kushya for the Shema), resolve
24 ms. Prolog 7 suites green (middot 2.9 s — pairwise scans over the
scroll). 5,853 verses loaded, 49,542 root occurrences.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>

talmudic_engine/src/reasoning/interpret.pl

@@ -111,16 +111,45 @@ find_relevant_mitzvot_traced(case(_Ref, Domain, Roots, _Question), Mitzvot, Coup
 find_relevant_verses(case(Ref, Domain, Roots, Question), Verses) :-
     find_relevant_verses_traced(case(Ref, Domain, Roots, Question), Verses, _).
 
-%% Traced version: returns verses AND the coupling trace for each
+%% Traced version: returns verses AND the coupling trace for each.
+%%
+%% Over the whole scroll (2026-10-19, l1_verses.pl) a common root lights
+%% hundreds of verses; a reading that cites them all cites nothing. The
+%% verses are ranked by how many of the case's roots they carry — a
+%% verse sharing three roots stands closer to the case than one sharing
+%% one — and the reading keeps the top ?MAX_VERSES, ties broken by
+%% canonical order. The coupling trace is kept only for the verses the
+%% reading cites: a trace names what the reading stands on.
+max_verses(12).
+
 find_relevant_verses_traced(case(_Ref, _Domain, Roots, _Question), Verses, Couplings) :-
-    findall(verse_ref(Book, Ch, Vs, Gloss)-coupled(root(Root), verse(Book, Ch, Vs)),
-            (   member(Root, Roots),
-                verse(Book, Ch, Vs, VerseRoots, Gloss),
+    sort(Roots, UniqueRoots),
+    findall(Shared-verse_ref(Book, Ch, Vs, Gloss),
+            (   verse(Book, Ch, Vs, VerseRoots, Gloss),
+                intersection(UniqueRoots, VerseRoots, SharedRoots),
+                SharedRoots \== [],
+                length(SharedRoots, N),
+                Shared is -N
+            ),
+            Scored),
+    %% keysort is stable: equal scores keep the scroll's order
+    keysort(Scored, Ranked),
+    pairs_values(Ranked, RankedVerses),
+    max_verses(Max),
+    take(Max, RankedVerses, Verses),
+    findall(coupled(root(Root), verse(Book, Ch, Vs)),
+            (   member(verse_ref(Book, Ch, Vs, _), Verses),
+                verse(Book, Ch, Vs, VerseRoots, _),
+                member(Root, UniqueRoots),
                 member(Root, VerseRoots)
             ),
-            AllPairs),
-    pairs_keys_values(AllPairs, AllVerses, Couplings),
-    sort(AllVerses, Verses).
+            Couplings).
+
+take(N, List, Prefix) :-
+    length(List, L),
+    (   L =< N -> Prefix = List
+    ;   length(Prefix, N), append(Prefix, _, List)
+    ).
 
 %% ============================================================
 %% CONSTRUCT INTERPRETATION