@rubytech/create-maxy 1.0.746 → 1.0.747

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@rubytech/create-maxy",
-  "version": "1.0.746",
+  "version": "1.0.747",
   "description": "Install Maxy — AI for Productive People",
   "bin": {
     "create-maxy": "./dist/index.js"

package/payload/platform/neo4j/schema.cypher

@@ -848,3 +848,43 @@ FOR (c:Credential) REQUIRE (c.accountId, c.name, c.authority) IS UNIQUE;
 
 CREATE INDEX credential_account IF NOT EXISTS
 FOR (c:Credential) ON (c.accountId);
+
+// ----------------------------------------------------------
+// Provenance indexes (Task 800) — back the shim's per-write
+// orphan check. Task 797's `executeWrite` runs an unanchored
+// `MATCH (n) WHERE n.createdBySession = $autoSession AND
+// n.createdAt >= $autoStartTimestamp AND NOT (n)--()` after
+// every operator write. Without an index this is `AllNodesScan`
+// — sub-millisecond on the current ~10K-node graph, but a
+// per-write tax that grows with graph size.
+//
+// Neo4j 5 property indexes require a label clause; there is no
+// label-free range/text/point index. The orphan check is
+// unanchored, so the planner relies on `UnionNodeByLabelsScan`
+// (Neo4j 5.6+) to combine per-label seeks. If a future EXPLAIN
+// shows `AllNodesScan` despite these indexes, the orphan-check
+// Cypher in `cypher-shim-write.ts` should be anchored with an
+// explicit label union — that's the structural fallback.
+//
+// Coverage: the five labels written most often through the raw
+// Cypher write path (Person, Organization, KnowledgeDocument,
+// Section, Task). Other written labels (Project, Position,
+// Credential, ToolCall, Conversation, etc.) still hit
+// `AllNodesScan` for the orphan check; extending coverage is a
+// separate decision, not a default.
+// ----------------------------------------------------------
+
+CREATE INDEX person_created_by_session IF NOT EXISTS
+FOR (n:Person) ON (n.createdBySession);
+
+CREATE INDEX organization_created_by_session IF NOT EXISTS
+FOR (n:Organization) ON (n.createdBySession);
+
+CREATE INDEX knowledge_doc_created_by_session IF NOT EXISTS
+FOR (n:KnowledgeDocument) ON (n.createdBySession);
+
+CREATE INDEX section_created_by_session IF NOT EXISTS
+FOR (n:Section) ON (n.createdBySession);
+
+CREATE INDEX task_created_by_session IF NOT EXISTS
+FOR (n:Task) ON (n.createdBySession);