feat: carve Jajce river into terrain

docs/BUILD_IN_PUBLIC_PLAN.md

@@ -774,17 +774,23 @@ Completed:
     stable-ID resource anchors in a resource-specific horizontal grid. Exact
     expanding queries preserve current scoring and selected targets while the
     reviewed 1,800-source case inspects about 4.4 candidates instead of 1,800.
+36. `Jajce River 03`: the deterministic Terrain3D source now carves a raised
+    waterfall shelf, upper stream, plunge basin, and curved downstream channel.
+    Shared watercourse profiles keep generated terrain and both water ribbons
+    aligned; softened foam/mist and a smaller rock frame make the falls part of
+    the landform. A matched Metal capture records the change, and the rebaked
+    navigation mesh passes terrain-height and runtime route checks.
 
 Next:
 
-1. Sculpt a real Terrain3D river channel and waterfall shelf before adding a
-   broad foliage asset pass. Use Terrain3D particles for nearby grass and
-   intentional instancing for tree/fern/rock masses, then rebake and validate
-   the required village/resource routes.
-2. During that first larger foliage pass, prove a bounded authoring workflow
-   that places intentional stable-ID `ResourceNode` anchors beside decorative
-   Terrain3D instances. Add livestock only with a real identity and interaction
-   contract.
+1. Prove one bounded foliage/resource authoring workflow at a riverbank or
+   forest edge. Use decorative Terrain3D/intentional instances for tree, fern,
+   flower, and rock masses, but create simulation-owned finite resources only
+   through selected stable-ID `ResourceNode` anchors.
+2. Validate that villagers discover both nearby and farther authored anchors,
+   preserve reachable interaction points, and keep the visual batch inside a
+   measured presentation budget. Add livestock only with a real identity and
+   interaction contract.
 
 Do not start with GIS data, a full city, a large asset pack, or more NPC
 mechanics. The next proof is a beautiful stage for the systems that already