Add enhanced leaderboard screen with data preparation

- Create screens/leaderboard.py with prepare_leaderboard_data function
- Add get_run_id_from_selection for navigation preparation
- Cache full leaderboard dataframe for run detail navigation
- Import and use leaderboard utilities in app.py
- Prepare foundation for Screen 1 → Screen 3 navigation

app.py

@@ -16,6 +16,7 @@ from utils.auth import is_authenticated, get_user_info, create_login_button, cre
 from utils.navigation import Navigator, Screen
 from data_loader import create_data_loader_from_env
 from mcp_client.sync_wrapper import get_sync_mcp_client
+from screens.leaderboard import prepare_leaderboard_data, get_run_id_from_selection
 
 # Initialize
 data_loader = create_data_loader_from_env()
@@ -24,10 +25,13 @@ mcp_client = get_sync_mcp_client()
 
 # Global state
 current_selected_run = None
+leaderboard_df_cache = None  # Cache full leaderboard with run_id column
 
 
 def load_leaderboard_view():
     """Load and display the leaderboard with MCP-powered insights"""
+    global leaderboard_df_cache
+
     # OAuth disabled for now
     # if not is_authenticated(token, profile):
     #     return "Please log in to view the leaderboard", ""
@@ -39,17 +43,11 @@ def load_leaderboard_view():
         if leaderboard_df.empty:
             return "No evaluation runs found in the leaderboard", ""
 
-        # Format dataframe for display
-        display_df = leaderboard_df[[
-            'model', 'agent_type', 'success_rate', 'total_tests',
-            'avg_duration_ms', 'total_cost_usd', 'co2_emissions_g'
-        ]].copy()
-
-        # Round numeric columns
-        display_df['success_rate'] = display_df['success_rate'].round(1)
-        display_df['avg_duration_ms'] = display_df['avg_duration_ms'].round(0)
-        display_df['total_cost_usd'] = display_df['total_cost_usd'].round(4)
-        display_df['co2_emissions_g'] = display_df['co2_emissions_g'].round(2)
+        # Cache the full dataframe (with run_id) for navigation
+        leaderboard_df_cache = leaderboard_df.copy()
+
+        # Prepare dataframe for display (formatted, sorted)
+        display_df = prepare_leaderboard_data(leaderboard_df)
 
         # Get MCP-powered insights
         try:

screens/__init__.py

@@ -0,0 +1 @@
+# Screens module

screens/leaderboard.py

@@ -0,0 +1,78 @@
+"""
+Leaderboard Screen for TraceMind-AI
+Displays evaluation runs with MCP-powered insights
+"""
+
+import pandas as pd
+import gradio as gr
+from typing import Optional, Tuple
+
+
+def prepare_leaderboard_data(df: pd.DataFrame) -> pd.DataFrame:
+    """
+    Prepare leaderboard dataframe for display
+
+    Args:
+        df: Raw leaderboard dataframe from HuggingFace
+
+    Returns:
+        Formatted dataframe for display
+    """
+    if df.empty:
+        return pd.DataFrame()
+
+    # Select and reorder columns for display
+    display_columns = [
+        'model', 'agent_type', 'provider', 'success_rate',
+        'total_tests', 'avg_duration_ms', 'total_cost_usd',
+        'co2_emissions_g', 'gpu_utilization_avg', 'submitted_by', 'timestamp'
+    ]
+
+    # Only include columns that exist
+    available_columns = [col for col in display_columns if col in df.columns]
+    display_df = df[available_columns].copy()
+
+    # Round numeric columns
+    if 'success_rate' in display_df.columns:
+        display_df['success_rate'] = display_df['success_rate'].round(1)
+    if 'avg_duration_ms' in display_df.columns:
+        display_df['avg_duration_ms'] = display_df['avg_duration_ms'].round(0)
+    if 'total_cost_usd' in display_df.columns:
+        display_df['total_cost_usd'] = display_df['total_cost_usd'].round(4)
+    if 'co2_emissions_g' in display_df.columns:
+        display_df['co2_emissions_g'] = display_df['co2_emissions_g'].round(2)
+    if 'gpu_utilization_avg' in display_df.columns:
+        display_df['gpu_utilization_avg'] = display_df['gpu_utilization_avg'].round(1)
+
+    # Sort by success rate descending by default
+    if 'success_rate' in display_df.columns:
+        display_df = display_df.sort_values('success_rate', ascending=False)
+
+    return display_df
+
+
+def get_run_id_from_selection(
+    df: pd.DataFrame,
+    evt: gr.SelectData
+) -> Optional[str]:
+    """
+    Extract run_id from a selected row in the dataframe
+
+    Args:
+        df: Full leaderboard dataframe (with run_id column)
+        evt: Gradio SelectData event from dataframe click
+
+    Returns:
+        run_id string or None
+    """
+    if df.empty or evt is None:
+        return None
+
+    try:
+        row_index = evt.index[0]  # evt.index is (row, col)
+        if row_index < len(df):
+            return df.iloc[row_index]['run_id']
+    except (IndexError, KeyError, AttributeError):
+        return None
+
+    return None