feat: Orderbook shadow logging system - Phase 1 complete

Implementation:
- Added 7 orderbook fields to Trade model (spreadBps, imbalanceRatio, depths, impact, walls)
- Oracle-based estimates with 2bps spread assumption
- ENV flag: ENABLE_ORDERBOOK_LOGGING (defaults true)
- Execute wrapper lines 1037-1053 guards orderbook logic

Database:
- Direct SQL ALTER TABLE (avoided migration drift issues)
- All columns nullable DOUBLE PRECISION
- Prisma schema synced via db pull + generate

Deployment:
- Container rebuilt and deployed successfully
- All 7 columns verified accessible
- System operational, ready for live trade validation

Files changed:
- config/trading.ts (enableOrderbookLogging flag, line 127)
- types/trading.ts (orderbook interfaces)
- lib/database/trades.ts (createTrade saves orderbook data)
- app/api/trading/execute/route.ts (ENV wrapper lines 1037-1053)
- prisma/schema.prisma (7 orderbook fields)
- docs/ORDERBOOK_SHADOW_LOGGING.md (complete documentation)

Status: ✅ PRODUCTION READY - awaiting first trade for validation

lib/database/trades.ts

@@ -55,6 +55,14 @@ export interface CreateTradeParams {
   pricePositionAtEntry?: number
   signalQualityScore?: number
   indicatorVersion?: string // TradingView Pine Script version (v5, v6, etc.)
+  // Orderbook metrics at entry (Phase 1 shadow logging - Dec 17, 2025)
+  obSpreadBps?: number           // Bid-ask spread in basis points
+  obImbalance?: number           // Buy vs sell pressure (-1 to +1, negative = more sellers)
+  obOppDepth0_2pctUSD?: number   // Opposing side liquidity within 0.2% in USD
+  obSameDepth0_2pctUSD?: number  // Same side liquidity within 0.2% in USD
+  obImpactBpsAtNotional?: number // Price impact for this trade size in basis points
+  obLargestOppWallBps?: number   // Distance to largest opposing wall in basis points
+  obLargestOppWallUSD?: number   // Size of largest opposing wall in USD
   // Phantom trade fields
   status?: string
   isPhantom?: boolean