feat: performance optimizations + part-materials validation

- @timed_step decorator with wall-clock + RSS tracking (pipeline_logger)
- Blender timing laps for sharp edges and material assignment
- MeshRegistry pattern: eliminate 13 scene.traverse() calls across viewers
- Lazy material cloning (clone-on-first-write in both viewers)
- _pipeline_session context manager: 7 create_engine() → 2 in render_thumbnail
- KD-tree spatial pre-filter for sharp edge marking (bbox-based pruning)
- Batch material library append: N bpy.ops.wm.append → single bpy.data.libraries.load
- GMSH single-session batching: compound all solids into one tessellation call
- Validate part-materials save endpoints against parsed_objects (prevents bogus keys)
- ROADMAP updated with completion status

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

render-worker/scripts/export_step_to_gltf.py

@@ -647,6 +647,9 @@ def main() -> None:
             )
 
             # Step 2: GMSH override for SOLID shapes (better seam topology)
+            # Batch all eligible solids into a single compound and tessellate in one
+            # GMSH session — avoids N × (gmsh init + brep write + brep read + finalize)
+            # overhead.  GMSH's internal OpenMP threading parallelizes across surfaces.
             _seen_shapes: list = []  # shapes already GMSH-tessellated; compared via IsSame()
 
             solids = []
@@ -661,6 +664,10 @@ def main() -> None:
                     solids.append(exp.Current())
                     exp.Next()
 
+            from OCP.TopoDS import TopoDS_Compound as _Compound
+            from OCP.BRep import BRep_Builder as _BBuilder
+
+            eligible = []
             for solid in solids:
                 # Skip REVERSED (mirrored) solids — keep BRepMesh tessellation.
                 # GMSH produces inverted-Jacobian meshes for negative-scale shapes.
@@ -673,9 +680,19 @@ def main() -> None:
                     continue
                 # Strip location: GMSH tessellates in definition space.
                 # The XCAF writer applies instance transforms at GLB export time.
-                solid_def = solid.Located(_TopLoc_Location())
-                _tessellate_with_gmsh(solid_def, args.linear_deflection, args.angular_deflection)
+                eligible.append(solid.Located(_TopLoc_Location()))
                 _seen_shapes.append(solid)
+
+            if eligible:
+                if len(eligible) == 1:
+                    _tessellate_with_gmsh(eligible[0], args.linear_deflection, args.angular_deflection)
+                else:
+                    compound = _Compound()
+                    bb = _BBuilder()
+                    bb.MakeCompound(compound)
+                    for s in eligible:
+                        bb.Add(compound, s)
+                    _tessellate_with_gmsh(compound, args.linear_deflection, args.angular_deflection)
     else:
         for i in range(1, free_labels.Length() + 1):
             shape = shape_tool.GetShape_s(free_labels.Value(i))