fix: close knowledge graph boundary gaps

backend/app/knowledge_graph.py

@@ -623,7 +623,7 @@ async def post_knowledge_graph(
           from post_summary_semantic_relationship
          where post_id = $1
          order by relation_ordinal
-         limit $2
+         limit ($2 + 1)
         """,
         post_id,
         relation_limit,
@@ -633,7 +633,7 @@ def semantic_key(node_type: str, name: str) -> str:
         digest = hashlib.sha256(f"{node_type}\0{name}".encode()).hexdigest()[:16]
         return f"semantic:{post_id}:{digest}"
 
-    for row in relation_rows:
+    for row in relation_rows[:relation_limit]:
         source = semantic_key(row["subject_type"], row["subject_name"])
         target = semantic_key(row["object_type"], row["object_name"])
         for key, node_type, name in (
@@ -670,7 +670,7 @@ def semantic_key(node_type: str, name: str) -> str:
         "post_id": post_id,
         "nodes": list(nodes.values()),
         "edges": edges,
-        "truncated": len(catalog_edges) > relation_limit or len(relation_rows) >= relation_limit,
+        "truncated": len(catalog_edges) > relation_limit or len(relation_rows) > relation_limit,
     }
 
 

backend/app/main.py

@@ -2645,12 +2645,13 @@ async def read_post_summary(
                     and load_settings().orchestrator_api_key
                 ),
             )
-            job = await ensure_post_content_job(
-                conn,
-                post_id,
-                raw_body,
-                content_complete=content_complete,
-            )
+            async with conn.transaction():
+                job = await ensure_post_content_job(
+                    conn,
+                    post_id,
+                    raw_body,
+                    content_complete=content_complete,
+                )
             if job.should_publish:
                 queue_event = (job.post_id, job.source_body_sha256)
             summary_waiting_for_images = not await post_content_summary_is_ready(conn, post_id)

docker/postgres-init/migrate.sh

@@ -18,7 +18,7 @@ for migration in /opt/lineageweave/migrations/*.sql; do
     migration_name=${migration##*/}
     case "$migration_name" in
         0012_*|0013_*|0014_*|0015_*|0016_*|0017_*|0018_*|0019_*|0020_*|0021_*|0022_*|0023_*|0024_*|0025_*|0026_*|0027_*|0028_*|0029_*|0030_*|0031_*|0032_*|0033_*|0034_*|0035_*|0036_*|0037_*|0038_*|0039_*|0040_*|0041_*|0042_*|0043_*|0044_*|0045_*|0046_*|0047_*|0048_*|0049_*|0050_*) ;;
-        0060_*|0100_*|0101_*|0102_*|0103_*|0104_*|0106_*|0107_*|0108_*|0109_*|0110_*|0111_*|0112_*|0113_*|0114_*) ;;
+        0060_*|0100_*|0101_*|0102_*|0103_*|0104_*|0105_*|0106_*|0107_*|0108_*|0109_*|0110_*|0111_*|0112_*|0113_*|0114_*) ;;
         *) continue ;;
     esac
     printf 'Applying %s\n' "$migration_name"

tests/test_knowledge_graph.py

@@ -16,6 +16,8 @@
 
 from __future__ import annotations
 
+import asyncio
+
 import pytest
 
 from lineageweave.knowledge_graph import (
@@ -30,6 +32,7 @@
     random_walk_with_restart,
     select_related_nodes,
 )
+from backend.app.knowledge_graph import post_knowledge_graph
 
 
 @pytest.fixture
@@ -72,6 +75,58 @@ def test_adaptive_depth_hub_reaches_more_nodes_than_a_sparse_node(synthetic_grap
     assert len(hub_related) > len(loner_related)
 
 
+@pytest.mark.parametrize(("overflow", "expected_truncated"), [(False, False), (True, True)])
+def test_post_knowledge_graph_relation_limit_boundary(
+    overflow: bool, expected_truncated: bool
+) -> None:
+    """A look-ahead row distinguishes an exact page from an overflow page."""
+    post_id = "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaa1"
+    organization_id = "bbbbbbbb-bbbb-bbbb-bbbb-bbbbbbbbbbb1"
+
+    class Connection:
+        def __init__(self) -> None:
+            self.semantic_query = ""
+            self.semantic_rows = [
+                {
+                    "relation_ordinal": 1,
+                    "subject_name": "Synthetic source",
+                    "subject_type": "organization",
+                    "predicate_code": "rel_voc",
+                    "object_name": "Synthetic customer",
+                    "object_type": "organization",
+                    "evidence_text": "Synthetic evidence",
+                    "relation_confidence": 0.9,
+                }
+            ]
+            if overflow:
+                self.semantic_rows.append({**self.semantic_rows[0], "relation_ordinal": 2})
+
+        async def fetch(self, query: str, *args: object) -> list[dict[str, object]]:
+            normalized = " ".join(query.lower().split())
+            if "select distinct person_id" in normalized:
+                return []
+            if "select distinct team_id" in normalized:
+                return []
+            if "select distinct corporate_entity_id" in normalized:
+                return [{"corporate_entity_id": organization_id}]
+            if "from knowledge_graph_edge edge" in normalized:
+                return []
+            if "from post_summary_semantic_relationship" in normalized:
+                self.semantic_query = normalized
+                return self.semantic_rows
+            return []
+
+        async def fetchval(self, query: str, *args: object) -> str | None:
+            return None
+
+    conn = Connection()
+    result = asyncio.run(post_knowledge_graph(conn, post_id, relation_limit=1))
+
+    assert "limit ($2 + 1)" in conn.semantic_query
+    assert result["truncated"] is expected_truncated
+    assert len(result["edges"]) == 1
+
+
 def test_start_node_absent_from_graph_returns_only_itself() -> None:
     scores = random_walk_with_restart({}, start_node="nowhere")
     assert scores == {"nowhere": 1.0}