diff --git a/Week-4/Task4.md b/Week-4/Task4.md
new file mode 100644
index 0000000..f33d856
--- /dev/null
+++ b/Week-4/Task4.md
@@ -0,0 +1,30 @@
+# Divya Week 4 Final Presentation: Visual Systems Capability Audit
+
+## 1. What I Explored In Weeks 1–3
+- **Material & Texture Systems**: Stress-tested the stability of the planetary variant switcher (Sun/Earth/Moon/Jupiter).
+- **Geometric Stability**: Benchmarked the robustness of primitive geometries (Sphere/Cube/Plane/Cylinder).
+- **Interaction Archetypes**: Built simulated configurator logic for product part-swapping and selection highlights.
+- **Edge Case Auditing**: Identified hard limits in the asset pipeline (5MB cap) and geometric renderer crashes.
+
+## 2. What Thob Does Well For Visual / Configurator Flows
+- **Zero-Latency Variant Switching**: The material variant engine is deterministic and visually flawless even under high stress.
+- **Professional Rendering Fidelity**: High-quality planetary mapping (UV integrity) is handled with professional-grade quality.
+- **Deterministic State Bridge**: Once logic is bound via $context, the relay between UI and 3D properties is 100% consistent (0% drift observed).
+
+## 3. Where Thob Is Weak / Fragile
+- **Lack of Native 3D Raycasting**: The inability to click the mesh itself forces a "Disconnected UX" that relies entirely on 2D UI overlays.
+- **Logic Scalability (State Explosion)**: Managing "Single-Select" behavior for 5+ objects is currently too manual and error-prone for professional use.
+- **Silent Asset Failures**: The lack of feedback in the asset pipeline (503/404 silent errors) creates a frustrating "Invisible Wall" for creators.
+
+## 4. My Most Important Recommendation
+- **The "Selection Group" Primitive**: We need a logic node that automates mutual exclusivity. This one feature would instantly move Thob from a "Point-Solution Toggler" to a "Professional Configurator Engine."
+
+## 5. My Best Quick Win
+- **Asset Pipeline "Error Toasts"**: Implement simple frontend validation for the 5MB file-size cap. Adding a single "File too large" notification would eliminate 80% of current authoring confusion.
+
+## 6. My Deeper Architecture Concern
+- **3D Interaction Engine**: Interaction shouldn't be an afterthought. Integrating a native Raycasting/Pointer event system into the Node tree is essential for Thob to feel like a modern, interactive 3D platform.
+
+## 7. If I Had 2 More Weeks
+- **Deep Component Hierarchy Benchmarking**: I would test how well Thob handles heavily nested structures (e.g., a car with 500+ sub-parts).
+- **Physics & Constraint Audit**: I'd explore if we can go beyond "swapping" to "moving" objects (exploding views) based on visual state triggers.