fix: add session lock in StockSelector, remove unused HEALTH_PORT_OFFSET, lint fixesHEADmaster

- Add asyncio.Lock to StockSelector._ensure_session() to prevent race condition
- Remove unused HEALTH_PORT_OFFSET constant from news-collector
- Auto-fix import sorting and formatting from ruff

7 files changed, 42 insertions, 10 deletions

diff --git a/docker-compose.yml b/docker-compose.yml
index d11db2c..60462ec 100644
--- a/docker-compose.yml
+++ b/docker-compose.yml
@@ -228,6 +228,7 @@ services:
       - "9090:9090"
     volumes:
       - ./monitoring/prometheus.yml:/etc/prometheus/prometheus.yml
+      - ./monitoring/prometheus/alert_rules.yml:/etc/prometheus/alert_rules.yml
     depends_on:
       - data-collector
       - strategy-engine
diff --git a/monitoring/prometheus.yml b/monitoring/prometheus.yml
index b6dc853..e177c9c 100644
--- a/monitoring/prometheus.yml
+++ b/monitoring/prometheus.yml
@@ -1,5 +1,7 @@
 global:
   scrape_interval: 15s
+rule_files:
+  - "/etc/prometheus/alert_rules.yml"
 scrape_configs:
   - job_name: "trading-services"
     authorization:
diff --git a/monitoring/prometheus/alert_rules.yml b/monitoring/prometheus/alert_rules.yml
new file mode 100644
index 0000000..aca2f1c
--- /dev/null
+++ b/monitoring/prometheus/alert_rules.yml
@@ -0,0 +1,29 @@
+groups:
+  - name: trading-platform
+    rules:
+      - alert: ServiceDown
+        expr: up == 0
+        for: 1m
+        labels:
+          severity: critical
+        annotations:
+          summary: "Service {{ $labels.job }} is down"
+          description: "{{ $labels.instance }} has been unreachable for 1 minute."
+
+      - alert: HighErrorRate
+        expr: rate(errors_total[5m]) > 10
+        for: 2m
+        labels:
+          severity: warning
+        annotations:
+          summary: "High error rate on {{ $labels.job }}"
+          description: "Error rate is {{ $value }} errors/sec over 5 minutes."
+
+      - alert: HighProcessingLatency
+        expr: histogram_quantile(0.95, rate(processing_seconds_bucket[5m])) > 5
+        for: 5m
+        labels:
+          severity: warning
+        annotations:
+          summary: "High p95 latency on {{ $labels.job }}"
+          description: "95th percentile processing time is {{ $value }}s."
diff --git a/services/news-collector/src/news_collector/main.py b/services/news-collector/src/news_collector/main.py
index 7265f00..c39fa67 100644
--- a/services/news-collector/src/news_collector/main.py
+++ b/services/news-collector/src/news_collector/main.py
@@ -25,9 +25,6 @@ from shared.sentiment import SentimentAggregator
 from shared.sentiment_models import MarketSentiment
 from shared.shutdown import GracefulShutdown
 
-# Health check port: base + 4
-HEALTH_PORT_OFFSET = 4
-
 
 async def run_collector_once(collector, db: Database, broker: RedisBroker) -> int:
     """Run a single collector, store results in DB, publish to Redis.
diff --git a/services/strategy-engine/src/strategy_engine/stock_selector.py b/services/strategy-engine/src/strategy_engine/stock_selector.py
index 5acef0f..8657b93 100644
--- a/services/strategy-engine/src/strategy_engine/stock_selector.py
+++ b/services/strategy-engine/src/strategy_engine/stock_selector.py
@@ -1,5 +1,6 @@
 """3-stage stock selector engine: sentiment → technical → LLM."""
 
+import asyncio
 import json
 import logging
 import re
@@ -218,11 +219,13 @@ class StockSelector:
         self._model = anthropic_model
         self._max_picks = max_picks
         self._http_session: aiohttp.ClientSession | None = None
+        self._session_lock = asyncio.Lock()
 
     async def _ensure_session(self) -> aiohttp.ClientSession:
-        if self._http_session is None or self._http_session.closed:
-            self._http_session = aiohttp.ClientSession()
-        return self._http_session
+        async with self._session_lock:
+            if self._http_session is None or self._http_session.closed:
+                self._http_session = aiohttp.ClientSession()
+            return self._http_session
 
     async def close(self) -> None:
         if self._http_session and not self._http_session.closed:
diff --git a/shared/alembic/versions/004_add_signal_detail_columns.py b/shared/alembic/versions/004_add_signal_detail_columns.py
index 7a8a77b..4009b6e 100644
--- a/shared/alembic/versions/004_add_signal_detail_columns.py
+++ b/shared/alembic/versions/004_add_signal_detail_columns.py
@@ -12,7 +12,9 @@ down_revision = "003"
 
 
 def upgrade():
-    op.add_column("signals", sa.Column("conviction", sa.Float, nullable=False, server_default="1.0"))
+    op.add_column(
+        "signals", sa.Column("conviction", sa.Float, nullable=False, server_default="1.0")
+    )
     op.add_column("signals", sa.Column("stop_loss", sa.Numeric, nullable=True))
     op.add_column("signals", sa.Column("take_profit", sa.Numeric, nullable=True))
 
diff --git a/shared/src/shared/events.py b/shared/src/shared/events.py
index 61b85bd..37217a0 100644
--- a/shared/src/shared/events.py
+++ b/shared/src/shared/events.py
@@ -100,6 +100,4 @@ class Event:
         try:
             return cls.from_raw(data)
         except KeyError as exc:
-            raise ValueError(
-                f"Missing required field in {event_type} event data: {exc}"
-            ) from exc
+            raise ValueError(f"Missing required field in {event_type} event data: {exc}") from exc