Add top 5 quality score calculation

data-science/analytics/quality_score.py

@@ -1,7 +1,19 @@
-"""Composite quality score — placeholder.
+"""Composite quality score calculation.
 
-TODO: Combine rating, reliability (error rate), usage, and review sentiment
-TODO: Output quality score 0-100 per agent
+This script reads mock agent usage and rating data, then calculates
+a simple quality score for each agent.
+
+Formula:
+quality_score = avg_rating * 20
+              + normalized_total_runs * 40
+              + reliability_score * 40
+
+Where:
+- avg_rating rewards agents with higher user ratings.
+- normalized_total_runs rewards agents that are used more often.
+- reliability_score = 1 - error_rate rewards stable agents with fewer errors.
+
+The script prints the top 5 agents by quality score.
 """
 
 import pandas as pd
@@ -14,18 +26,23 @@ def compute_quality_scores() -> pd.DataFrame:
     usage = pd.read_csv(DATASETS_DIR / "agent_usage.csv")
     ratings = pd.read_csv(DATASETS_DIR / "agent_ratings.csv")
     merged = usage.merge(ratings, on="agent_id", how="left")
-
-    # TODO: Weighted composite formula
+
+    if "agent_name" not in merged.columns:
+   	 merged["agent_name"] = merged["agent_id"]	
+
     merged["quality_score"] = (
         merged["avg_rating"] * 20
         + (merged["total_runs"] / merged["total_runs"].max()) * 40
         + (1 - merged["error_rate"]) * 40
     ).round(1)
 
-    return merged[["agent_id", "agent_name", "quality_score"]].sort_values(
-        "quality_score", ascending=False
+    return (
+	 merged[["agent_id", "agent_name", "quality_score"]]
+	.sort_values("quality_score", ascending=False)
+	.head(5)
     )
 
 
 if __name__ == "__main__":
+    print("Top 5 Agents by Quality Score:")
     print(compute_quality_scores().to_string(index=False))