feat: author resource-aware riverbank grove

docs/LEARNING_ROADMAP.md

@@ -896,17 +896,30 @@ plunge basin, and curved downstream channel. A matched Metal capture records
 the result, and the Terrain3D-derived navigation mesh was rebaked and validated
 against required village/resource routes.
 
-The immediate next slice should prove one bounded foliage/resource authoring
-workflow in the real world: build a restrained riverbank or forest-edge cluster
-with decorative Terrain3D/intentional instances, pair only selected harvestable
-trees or berry patches with stable-ID `ResourceNode` anchors, and verify nearby
-and farther loaded anchors remain discoverable through `ActiveWorldAdapter`.
-Keep decorative instances out of simulation state, measure the bounded visual
-cost, and do not generalize the resource grid to people/buildings/events or
+The first authored foliage/resource workflow is complete. One riverbank cluster
+uses three intentional MultiMesh batches for 28 fern, flower, and stone
+instances plus two decorative trees, while only an existing stable-ID berry and
+tree are finite `ResourceNode` anchors. Both anchors preserve the save contract
+of eighteen resources, sit above the carved channel, and expose separate
+terrain-snapped approach points. A forced village query expands beyond the
+initial 24 m range and selects the nested berry exactly by ID.
+
+The immediate next slice should make those two resource-specific visuals
+reflect authoritative amount/depletion: thin the berry patch and leave a stump
+or equivalent readable tree state after real NPC/player extraction, then prove
+save/restore rebuilds the same presentation without persisting visual-only
+state. Only after both the berry and tree consume that contract should shared
+cluster/depletion tooling be extracted or the workflow copied to a forest
+edge. Do not generalize the resource grid to people/buildings/events or
 introduce active/abstract LOD before those consumers need it.
 
 Recently completed:
 
+- `Jajce Resource Grove 04`: two existing river resources now live as nested
+  stable-ID anchors inside a restrained decorative bank cluster. Twenty-eight
+  understory instances use three MultiMesh batches, bespoke berry/tree visuals
+  replace generic boxes, terrain-snapped approach points remain reachable, and
+  a real far-source query proves parent-path-independent expanding discovery.
 - `Jajce River 03`: deterministic Terrain3D generation now carves the upper
   stream, raised waterfall shelf, plunge basin, and curved downstream channel;
   shared watercourse geometry keeps both water ribbons aligned, soft masked