fix(dedup): a rule or preference create surfaces what it overlaps by meaning (#4134)

find_duplicate_rule was title-only, on the stated premise that rules are not
a semantic-retrieval surface - false since rules were embedded. A preference
restating a rule under another title passed untouched, and since both kinds
share one ranking, the weaker label could arrive alone.

find_overlapping_rules queries semantic_search_rules with the rule_document
shape, both kinds, in the scope the new record ranks in (global: every rule
the caller owns; project: global + that project). All three MCP create doors
call it before creating and return overlaps + overlap_note on the reply.

It advises rather than blocks, on measurement: across 16 sampled records the
nearest DISTINCT neighbour reached 0.853, while a true rewording scored 0.850.
No threshold separates the bands, so the floor (0.80) sits below the
restatement and the author judges. The stale docstring is corrected.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01821k5B3Ysecp9fNYs92Kuy

src/scribe/mcp/tools/rulebooks.py

@@ -489,13 +489,20 @@ async def create_rule(
         order_index: Display order within the topic (default 0).
         force: Bypass the near-duplicate gate. By default, a title-identical rule
             already in this topic BLOCKS creation and returns its id so you update
-            it instead. Set true only for a genuinely distinct rule.
+            it instead. Set true only for a genuinely distinct rule. A rule or
+            preference that answers the same MOMENT under another title does
+            not block: the create goes through and the reply carries
+            `overlaps` and `overlap_note` — read the top one and decide.
     """
     uid = current_user_id()
     if not force:
         dup = await dedup_svc.find_duplicate_rule(title, topic_id=topic_id)
         if dup is not None:
             return dedup_svc.duplicate_response(dup, "rule")
+    # Before the create, so the new rule cannot find itself (#4134).
+    overlaps = await dedup_svc.find_overlapping_rules(
+        uid, title, statement, when_to_apply,
+    )
     rule = await rulebooks_svc.create_rule(
         topic_id=topic_id, user_id=uid,
         title=title, statement=statement, when_to_apply=when_to_apply,
@@ -503,7 +510,9 @@ async def create_rule(
         why=why, how_to_apply=how_to_apply, order_index=order_index,
         verify_with=verify_with, expires_when=expires_when,
     )
-    return await rulebooks_svc.rule_detail(uid, rule, system_ids)
+    data = await rulebooks_svc.rule_detail(uid, rule, system_ids)
+    data.update(dedup_svc.overlap_response(overlaps, "rule"))
+    return data
 
 
 async def create_project_rule(
@@ -582,7 +591,9 @@ async def create_project_rule(
         order_index: Display order within the project's rule list (default 0).
         force: Bypass the near-duplicate gate. By default, a title-identical rule
             already on this project BLOCKS creation and returns its id so you
-            update it instead. Set true only for a genuinely distinct rule.
+            update it instead. Set true only for a genuinely distinct rule. An
+            overlap by meaning never blocks; it arrives as `overlaps` and
+            `overlap_note` on the reply — see create_rule.
     """
     uid = current_user_id()
     derived_title = title.strip() or statement.strip().split(".")[0][:50]
@@ -590,6 +601,9 @@ async def create_project_rule(
         dup = await dedup_svc.find_duplicate_rule(derived_title, project_id=project_id)
         if dup is not None:
             return dedup_svc.duplicate_response(dup, "rule")
+    overlaps = await dedup_svc.find_overlapping_rules(
+        uid, derived_title, statement, when_to_apply, project_id=project_id,
+    )
     rule = await rulebooks_svc.create_project_rule(
         project_id=project_id, user_id=uid,
         title=derived_title, statement=statement, when_to_apply=when_to_apply,
@@ -597,7 +611,9 @@ async def create_project_rule(
         why=why, how_to_apply=how_to_apply, order_index=order_index,
         verify_with=verify_with, expires_when=expires_when,
     )
-    return await rulebooks_svc.rule_detail(uid, rule, system_ids)
+    data = await rulebooks_svc.rule_detail(uid, rule, system_ids)
+    data.update(dedup_svc.overlap_response(overlaps, "rule"))
+    return data
 
 
 async def update_rule(
@@ -784,7 +800,10 @@ async def create_preference(
             a preference could only be filed after the fact (#4249).
         force: Bypass the near-duplicate gate. For a genuinely distinct
             preference, not for one that is "mostly" different — a mostly
-            different preference is an update.
+            different preference is an update. A RULE that already answers
+            this moment comes back as `overlaps` / `overlap_note` on the reply
+            rather than blocking; if it says the same thing, the preference is
+            the weaker copy of it and should go.
     """
     uid = current_user_id()
     if not when_to_apply.strip():
@@ -803,13 +822,18 @@ async def create_preference(
         dup = await dedup_svc.find_duplicate_rule(title, topic_id=topic_id)
         if dup is not None:
             return dedup_svc.duplicate_response(dup, "rule")
+    overlaps = await dedup_svc.find_overlapping_rules(
+        uid, title, statement, when_to_apply,
+    )
     rule = await rulebooks_svc.create_rule(
         topic_id=topic_id, user_id=uid,
         title=title, statement=statement, when_to_apply=when_to_apply,
         kind="preference", arose_from_id=arose_from_id,
         why=why, how_to_apply=how_to_apply, order_index=order_index,
     )
-    return await rulebooks_svc.rule_detail(uid, rule, system_ids)
+    data = await rulebooks_svc.rule_detail(uid, rule, system_ids)
+    data.update(dedup_svc.overlap_response(overlaps, "preference"))
+    return data
 
 
 async def update_preference(