Mark TX1→NC1 migration as NOT RECOMMENDED (analysis shows no benefit)

docs/core/tasks.md

@@ -714,15 +714,28 @@ docker exec ollama ollama pull llama3.2-vision:11b # Images (8GB RAM)
 
 ---
 
-### **Optional: Migrate TX1 Games to NC1**
+### **⚠️ NOT RECOMMENDED: Migrate TX1 Games to NC1**
 
-**TX1 game servers:** 24GB total  
-**NC1 after cleanup:** 46GB free  
-**Result:** NC1 at 73% usage (healthy)
+**Analysis (Feb 14, 2026):** Migration provides NO meaningful benefit and introduces unnecessary risk.
 
-**Can defer this** — AI and games can coexist on TX1
+**Why NOT to migrate:**
+- ❌ No performance gain (identical hardware)
+- ❌ 3-6 hours work for purely aesthetic benefit
+- ❌ Loses geographic/hardware redundancy (currently TX1 Dallas + NC1 Charlotte)
+- ❌ Single point of failure (all 15 servers on NC1)
+- ❌ Storage constraints (NC1: 98GB total vs TX1: 911GB total)
+- ❌ Player disruption (new IPs, Discord updates needed)
 
-**If migrating:** 30-60 min per server via Pterodactyl transfer
+**Why TX1 + AI coexistence works:**
+- ✅ TX1 has 809GB free storage (AI needs ~100GB, games use 24GB = 685GB still free)
+- ✅ TX1 has 222GB free RAM (AI needs ~90GB, games use 20GB = 112GB still free)
+- ✅ Separate directories (/opt/ai-stack vs /var/lib/pterodactyl)
+- ✅ No resource conflicts
+- ✅ 14% storage usage, 50% RAM usage (very comfortable)
+
+**Conclusion:** Deploy AI on TX1 alongside existing games. Leave infrastructure as-is.
+
+**Only migrate if:** TX1 storage/RAM becomes constrained (extremely unlikely with current headroom)
 
 ---