diff --git a/sql/astock_kg_logic_views.sql b/sql/astock_kg_logic_views.sql
index 746764b..d8dc05e 100644
--- a/sql/astock_kg_logic_views.sql
+++ b/sql/astock_kg_logic_views.sql
@@ -223,7 +223,14 @@ SELECT CASE
        r.beneficiary                                             AS subject_name,
        CASE WHEN r.cluster_key LIKE 'Company:%'
             THEN substr(r.cluster_key, length('Company:') + 1) END AS ts_code,
-       CASE WHEN r.cluster_key LIKE 'Segment:%'
+       -- 双冒号那一支必须写在前面：双冒号的键同样满足单冒号的前缀匹配，
+       -- 顺序反了会截出一个多余的冒号。
+       -- 双冒号 Segment:: 是 2026-09-04 起按投影环节做的评析，环节名与传导目标同一命名空间；
+       -- 单冒号 Segment:  是逐条论断聚簇出来的研报自由短语，在投影里未必查得到这个环节。
+       -- scope 的判定不用改：双冒号同样满足 LIKE 'Segment:%'，两者都归 segment。
+       CASE WHEN r.cluster_key LIKE 'Segment::%'
+            THEN substr(r.cluster_key, length('Segment::') + 1)
+            WHEN r.cluster_key LIKE 'Segment:%'
             THEN substr(r.cluster_key, length('Segment:') + 1) END AS segment_name,
        r.comparison->>'leaning'                                  AS leaning,
        r.comparison->'core_contradiction'->>'text'               AS core_contradiction,