feat: aggregate all sentiment sources into composite score

Redesign /stock/{symbol}/sentiment to combine 4 data sources with
weighted scoring:
- News sentiment (Alpha Vantage, 25%) - article-level bullish/bearish
- Analyst recommendations (Finnhub, 30%) - buy/sell ratio
- Upgrade/downgrade activity (yfinance, 20%) - recent actions
- Reddit buzz (ApeWisdom, 25%) - mention change trend

Returns composite_score (-1 to +1), composite_label, per-source
scores, and full detail data from each source.

routes_sentiment.py

@@ -23,22 +23,112 @@ router = APIRouter(prefix="/api/v1")
 @router.get("/stock/{symbol}/sentiment", response_model=ApiResponse)
 @safe
 async def stock_sentiment(symbol: str = Path(..., min_length=1, max_length=20)):
-    """Get aggregated sentiment: Alpha Vantage news sentiment + Finnhub analyst data."""
+    """Aggregated sentiment from all sources with composite score.
+
+    Combines: Alpha Vantage news sentiment, Finnhub analyst data,
+    Reddit mentions, and analyst upgrades into a single composite score.
+    Score range: -1.0 (extreme bearish) to +1.0 (extreme bullish).
+    """
     symbol = validate_symbol(symbol)
-    finnhub_data, av_data = await asyncio.gather(
-        finnhub_service.get_sentiment_summary(symbol),
+
+    # Fetch all sources in parallel
+    av_data, finnhub_data, reddit_data, upgrades_data, recs_data = await asyncio.gather(
         alphavantage_service.get_news_sentiment(symbol, limit=20),
+        finnhub_service.get_sentiment_summary(symbol),
+        finnhub_service.get_reddit_sentiment(symbol),
+        openbb_service.get_upgrades_downgrades(symbol, limit=10),
+        finnhub_service.get_recommendation_trends(symbol),
         return_exceptions=True,
     )
-    if isinstance(finnhub_data, BaseException):
-        logger.exception("Finnhub error", exc_info=finnhub_data)
-        finnhub_data = {}
-    if isinstance(av_data, BaseException):
-        logger.exception("Alpha Vantage error", exc_info=av_data)
-        av_data = {}
 
-    data = {**finnhub_data, "alpha_vantage_sentiment": av_data}
-    return ApiResponse(data=data)
+    def _safe(result, default):
+        return default if isinstance(result, BaseException) else result
+
+    av_data = _safe(av_data, {})
+    finnhub_data = _safe(finnhub_data, {})
+    reddit_data = _safe(reddit_data, {})
+    upgrades_data = _safe(upgrades_data, [])
+    recs_data = _safe(recs_data, [])
+
+    # --- Score each source ---
+    scores: list[tuple[str, float, float]] = []  # (source, score, weight)
+
+    # 1. News sentiment (Alpha Vantage): avg_score ranges ~-0.35 to +0.35
+    if isinstance(av_data, dict) and av_data.get("overall_sentiment"):
+        av_score = av_data["overall_sentiment"].get("avg_score")
+        if av_score is not None:
+            # Normalize to -1..+1 (AV scores are typically -0.35 to +0.35)
+            normalized = max(-1.0, min(1.0, av_score * 2.5))
+            scores.append(("news", round(normalized, 3), 0.25))
+
+    # 2. Analyst recommendations (Finnhub): buy/sell ratio
+    if isinstance(recs_data, list) and recs_data:
+        latest = recs_data[0]
+        total = sum(latest.get(k, 0) for k in ("strongBuy", "buy", "hold", "sell", "strongSell"))
+        if total > 0:
+            bullish = latest.get("strongBuy", 0) + latest.get("buy", 0)
+            bearish = latest.get("sell", 0) + latest.get("strongSell", 0)
+            ratio = (bullish - bearish) / total  # -1 to +1
+            scores.append(("analysts", round(ratio, 3), 0.30))
+
+    # 3. Analyst upgrades vs downgrades (yfinance)
+    if isinstance(upgrades_data, list) and upgrades_data:
+        ups = sum(1 for u in upgrades_data if u.get("action") in ("up", "init"))
+        downs = sum(1 for u in upgrades_data if u.get("action") == "down")
+        maintains = len(upgrades_data) - ups - downs
+        if len(upgrades_data) > 0:
+            upgrade_score = (ups - downs) / len(upgrades_data)
+            scores.append(("upgrades", round(upgrade_score, 3), 0.20))
+
+    # 4. Reddit buzz (ApeWisdom)
+    if isinstance(reddit_data, dict) and reddit_data.get("found"):
+        mentions = reddit_data.get("mentions_24h", 0)
+        change = reddit_data.get("mentions_change_pct")
+        if change is not None and mentions > 10:
+            # Positive change = bullish buzz, capped at +/- 1
+            reddit_score = max(-1.0, min(1.0, change / 100))
+            scores.append(("reddit", round(reddit_score, 3), 0.25))
+
+    # --- Compute weighted composite ---
+    if scores:
+        total_weight = sum(w for _, _, w in scores)
+        composite = sum(s * w for _, s, w in scores) / total_weight
+        composite = round(composite, 3)
+    else:
+        composite = None
+
+    # Map to label
+    if composite is None:
+        label = "Unknown"
+    elif composite >= 0.5:
+        label = "Strong Bullish"
+    elif composite >= 0.15:
+        label = "Bullish"
+    elif composite >= -0.15:
+        label = "Neutral"
+    elif composite >= -0.5:
+        label = "Bearish"
+    else:
+        label = "Strong Bearish"
+
+    return ApiResponse(data={
+        "symbol": symbol,
+        "composite_score": composite,
+        "composite_label": label,
+        "source_scores": {name: score for name, score, _ in scores},
+        "source_weights": {name: weight for name, _, weight in scores},
+        "details": {
+            "news_sentiment": av_data if isinstance(av_data, dict) else {},
+            "analyst_recommendations": recs_data[0] if isinstance(recs_data, list) and recs_data else {},
+            "recent_upgrades": upgrades_data[:5] if isinstance(upgrades_data, list) else [],
+            "reddit": reddit_data if isinstance(reddit_data, dict) else {},
+            "finnhub_news": (
+                finnhub_data.get("recent_news", [])[:5]
+                if isinstance(finnhub_data, dict)
+                else []
+            ),
+        },
+    })
 
 
 @router.get("/stock/{symbol}/news-sentiment", response_model=ApiResponse)