Add --publish flag for Render publish-only deployment

render.yaml

@@ -0,0 +1,27 @@
+services:
+  - type: worker
+    name: allocation-engine-2.0
+    runtime: python
+    region: oregon
+    plan: starter
+    buildCommand: pip install -r requirements.txt
+    startCommand: python -m trading_system.main --publish --dashboard --continuous --interval 5 --ticker BTC SPY QQQ AMZN
+    envVars:
+      - key: NETLIFY_API_TOKEN
+        sync: false
+      - key: NETLIFY_SITE_ID
+        sync: false
+      - key: TWELVE_DATA_API_KEY
+        sync: false
+      - key: RH_USERNAME
+        sync: false
+      - key: RH_PASSWORD
+        sync: false
+      - key: RH_TOTP_CODE
+        sync: false
+      - key: RH_AUTOMATED_ACCOUNT_NUMBER
+        sync: false
+      - key: SLACK_WEBHOOK_URL
+        sync: false
+      - key: PYTHON_VERSION
+        value: "3.11"

trading_system/main.py

@@ -35,7 +35,8 @@ def __init__(self, twelve_data_api_key: str, symbols: List[str],
                  strategy_name: str = 'momentum_dca',
                  verbose: bool = False,
                  dashboard: bool = False,
-                 recent_days: int = None):
+                 recent_days: int = None,
+                 publish: bool = False):
         """
         Initialize trading system
 
@@ -48,50 +49,58 @@ def __init__(self, twelve_data_api_key: str, symbols: List[str],
             verbose: If True, show detailed output (metrics, portfolio, etc.)
             dashboard: If True, fetch market indicators and write dashboard data each cycle
             recent_days: If set, limit backtest to last N daily bars
+            publish: If True, upload blobs even in dry-run mode (no trades, but site data updates)
         """
         self.symbols = symbols
         self.dry_run = dry_run
         self.strategy_name = strategy_name
         self.verbose = verbose
         self.dashboard = dashboard
         self.recent_days = recent_days
+        self.publish = publish
 
         # Initialize components
         self.data_provider = TwelveDataProvider(twelve_data_api_key)
         self.metrics_calculator = MetricsCalculator()
         self.state_manager = StateManager()
-        self.trading_bot = SafeCashBot()
 
-        # Initialize execution quality layer
+        # Skip broker initialization in publish-only mode (no trading needed)
+        self.trading_bot = None
         self.fill_logger = None
-        try:
-            from trading_system.execution.fill_auditor import FillAuditor
-            from trading_system.execution.spread_checker import SpreadChecker
-            from trading_system.execution.price_optimizer import PriceOptimizer
-            from trading_system.execution.pdt_gate import PDTGate
-            from trading_system.execution.fill_logger import FillLogger
-
-            fill_auditor = FillAuditor(
-                alpaca_key=os.getenv('ALPACA_API_KEY', ''),
-                alpaca_secret=os.getenv('ALPACA_SECRET_KEY', ''),
-                twelve_data_provider=self.data_provider,
-            )
-            spread_checker = SpreadChecker(fill_auditor=fill_auditor)
-            price_optimizer = PriceOptimizer()
-            pdt_gate = PDTGate(trading_bot=self.trading_bot)
-            fill_logger = FillLogger()
-
-            self.trading_bot.init_execution_layer(
-                fill_auditor=fill_auditor,
-                spread_checker=spread_checker,
-                price_optimizer=price_optimizer,
-                pdt_gate=pdt_gate,
-                fill_logger=fill_logger,
-            )
+        if not (self.publish and self.dry_run):
+            self.trading_bot = SafeCashBot()
 
-            self.fill_logger = fill_logger
-        except Exception as e:
-            print(f"  [exec-layer] Execution quality layer init failed (proceeding without): {e}")
+            # Initialize execution quality layer
+            try:
+                from trading_system.execution.fill_auditor import FillAuditor
+                from trading_system.execution.spread_checker import SpreadChecker
+                from trading_system.execution.price_optimizer import PriceOptimizer
+                from trading_system.execution.pdt_gate import PDTGate
+                from trading_system.execution.fill_logger import FillLogger
+
+                fill_auditor = FillAuditor(
+                    alpaca_key=os.getenv('ALPACA_API_KEY', ''),
+                    alpaca_secret=os.getenv('ALPACA_SECRET_KEY', ''),
+                    twelve_data_provider=self.data_provider,
+                )
+                spread_checker = SpreadChecker(fill_auditor=fill_auditor)
+                price_optimizer = PriceOptimizer()
+                pdt_gate = PDTGate(trading_bot=self.trading_bot)
+                fill_logger = FillLogger()
+
+                self.trading_bot.init_execution_layer(
+                    fill_auditor=fill_auditor,
+                    spread_checker=spread_checker,
+                    price_optimizer=price_optimizer,
+                    pdt_gate=pdt_gate,
+                    fill_logger=fill_logger,
+                )
+
+                self.fill_logger = fill_logger
+            except Exception as e:
+                print(f"  [exec-layer] Execution quality layer init failed (proceeding without): {e}")
+        else:
+            print("  [publish-only] Skipping broker init — publish-only mode (no trades)")
 
         if strategy_name == 'momentum_dca_long':
             self.strategy = MomentumDcaLongStrategy(symbols)
@@ -607,15 +616,18 @@ def run_once(self):
             print(f"{'='*70}\n")
 
             # Print initial portfolio allocation
-            self.print_portfolio_allocation()
+            if self.trading_bot:
+                self.print_portfolio_allocation()
 
         # Fetch open orders once (used by momentum_dca)
         open_orders = []
-        if self.strategy_name == 'momentum_dca_long':
-            open_orders = self.trading_bot.get_open_orders()
-
-        recent_orders = self.trading_bot.get_recent_orders(days=7)
-        recent_option_orders = self.trading_bot.get_recent_option_orders(days=7)
+        recent_orders = []
+        recent_option_orders = []
+        if self.trading_bot:
+            if self.strategy_name == 'momentum_dca_long':
+                open_orders = self.trading_bot.get_open_orders()
+            recent_orders = self.trading_bot.get_recent_orders(days=7)
+            recent_option_orders = self.trading_bot.get_recent_option_orders(days=7)
 
         # Print order book before processing through state manager
         if open_orders:
@@ -668,11 +680,12 @@ def run_once(self):
                 if self.verbose:
                     print(self.metrics_calculator.format_metrics(symbol, metrics))
 
-                # 3. Execute strategy
-                signal = self.execute_strategy(symbol, metrics, open_orders)
+                # 3. Execute strategy (requires broker for positions)
+                if self.trading_bot:
+                    signal = self.execute_strategy(symbol, metrics, open_orders)
 
-                # 4. Process signal
-                self.process_signal(symbol, signal, open_orders)
+                    # 4. Process signal
+                    self.process_signal(symbol, signal, open_orders)
 
             except Exception as e:
                 print(f"Error processing {symbol}: {e}")
@@ -683,14 +696,14 @@ def run_once(self):
 
         # Log state: local file in dry-run, Netlify Blobs when live
         symbols_filter = None if self.verbose else self.symbols
-        portfolio_data = self.trading_bot.get_portfolio_summary(symbols=symbols_filter)
+        portfolio_data = self.trading_bot.get_portfolio_summary(symbols=symbols_filter) if self.trading_bot else None
 
         # Compute drift metrics from cached daily bars (if available)
         drift_metrics = self._compute_drift_metrics()
 
         log_state_to_blob(
             self.state_manager,
-            live=not self.dry_run,
+            live=not self.dry_run or self.publish,
             order_book=open_orders,
             portfolio=portfolio_data,
             drift_metrics=drift_metrics,
@@ -737,8 +750,9 @@ def run_once(self):
             self.state_manager.print_state_summary()
 
             # Print final portfolio allocation
-            print("\nFinal Portfolio Allocation:")
-            self.print_portfolio_allocation()
+            if self.trading_bot:
+                print("\nFinal Portfolio Allocation:")
+                self.print_portfolio_allocation()
 
         # Print fill quality stats
         if hasattr(self, 'fill_logger') and self.fill_logger:
@@ -848,7 +862,7 @@ def _send_oncall_summary(self, open_orders: List[Dict], portfolio_data: Dict):
             lines.append("=" * 40)
 
             # PDT status
-            pdt_info = self.trading_bot.get_pdt_status()
+            pdt_info = self.trading_bot.get_pdt_status() if self.trading_bot else None
             if pdt_info is not None:
                 count = pdt_info.get('day_trade_count', 0)
                 if pdt_info.get('flagged'):
@@ -1137,6 +1151,11 @@ def main():
         action='store_true',
         help='Fetch market indicators and write dashboard/market_data.json each cycle'
     )
+    parser.add_argument(
+        '--publish',
+        action='store_true',
+        help='Upload blobs in dry-run mode (no trades placed, but site data updates)'
+    )
     parser.add_argument(
         '--backtest',
         action='store_true',
@@ -1179,10 +1198,11 @@ def main():
         verbose=args.verbose,
         dashboard=args.dashboard,
         recent_days=args.recent,
+        publish=args.publish,
     )
 
     # Notify Slack on deployment startup
-    mode = "LIVE" if args.live else "DRY RUN"
+    mode = "LIVE" if args.live else ("PUBLISH" if args.publish else "DRY RUN")
     send_slack_alert(
         f":rocket: Engine deployed — mode={mode}, strategy={args.strategy}, symbols={', '.join(symbols)}",
         emoji=":rocket:",