Update docker/entrypoint-hf.sh

docker/entrypoint-hf.sh

@@ -1,85 +1,130 @@
-# Dockerfile for Hugging Face Spaces - Gradio Multi-Agent Assistant with MCP Toolbox
-# This deploys the complete multi-agent system with Cloud SQL access via MCP Toolbox
-
-FROM python:3.11-slim
-
-# Install system dependencies including Node.js
-RUN apt-get update && apt-get install -y \
-    ca-certificates \
-    curl \
-    git \
-    build-essential \
-    gettext-base \
-    && curl -fsSL https://deb.nodesource.com/setup_20.x | bash - \
-    && apt-get install -y nodejs \
-    && rm -rf /var/lib/apt/lists/*
-
-# Install MCP Toolbox binary (v0.18.0)
-ARG TOOLBOX_VERSION=v0.18.0
-RUN ARCH=$(dpkg --print-architecture) && \
-    if [ "$ARCH" = "amd64" ]; then \
-        TOOLBOX_ARCH="amd64"; \
-    elif [ "$ARCH" = "arm64" ]; then \
-        TOOLBOX_ARCH="arm64"; \
-    else \
-        echo "Unsupported architecture: $ARCH" && exit 1; \
-    fi && \
-    curl -L "https://storage.googleapis.com/genai-toolbox/${TOOLBOX_VERSION}/linux/${TOOLBOX_ARCH}/toolbox" \
-        -o /usr/local/bin/toolbox && \
-    chmod +x /usr/local/bin/toolbox
-
-# Verify installations
-RUN python --version && node --version && npm --version && toolbox --version
-
-# Set working directory
-WORKDIR /app
-
-# Copy requirements first (for layer caching)
-COPY requirements.txt .
-
-# Install Python dependencies
-RUN pip install --no-cache-dir --upgrade pip && \
-    pip install --no-cache-dir -r requirements.txt
-
-# Install npx-based MCP servers globally
-# CoinGecko MCP server for crypto data
-RUN npm install -g @coingecko/coingecko-mcp
-
-# Install uvx for Python-based MCP servers
-RUN pip install --no-cache-dir uv
-
-# Copy application code
-COPY src/ ./src/
-COPY ui/ ./ui/
-
-# Make src a proper Python package
-RUN find /app/src -type d -exec touch {}/__init__.py \;
-
-# Copy MCP Toolbox configuration
-COPY tools.yaml /app/tools.yaml
-
-# Create directory for temporary file uploads
-RUN mkdir -p /tmp/uploads && chmod 777 /tmp/uploads
-
-# Copy entrypoint script
-COPY docker/entrypoint-hf.sh /app/entrypoint.sh
-RUN chmod +x /app/entrypoint.sh
-
-# Set environment variables for Hugging Face Spaces
-ENV GRADIO_SERVER_NAME="0.0.0.0" \
-    GRADIO_SERVER_PORT="7860" \
-    PYTHONUNBUFFERED=1 \
-    PYTHONPATH=/app
-
-# Expose Gradio port (HF Spaces expects 7860)
-EXPOSE 7860
-
-# Expose MCP Toolbox port (internal - for communication between services)
-EXPOSE 5000
-
-# Health check
-HEALTHCHECK --interval=30s --timeout=10s --start-period=90s --retries=3 \
-    CMD curl -f http://localhost:7860/ || exit 1
-
-# Run Gradio app via entrypoint
-CMD ["/app/entrypoint.sh"]
+#!/bin/bash
+# Entrypoint script for Hugging Face Spaces deployment
+# Launches MCP Toolbox and Gradio multi-agent assistant
+
+set -e
+
+echo "=================================================="
+echo "🚀 Starting PortfolioMind Multi-Agent Assistant"
+echo "=================================================="
+
+# Validate required environment variables
+echo "📋 Validating configuration..."
+
+REQUIRED_VARS=(
+    "GOOGLE_API_KEY"
+    "CHROMA_API_KEY"
+    "CHROMA_TENANT"
+    "CHROMA_DATABASE"
+    "ALPHA_VANTAGE_API_KEY"
+)
+
+MISSING_VARS=()
+for var in "${REQUIRED_VARS[@]}"; do
+    if [ -z "${!var}" ]; then
+        MISSING_VARS+=("$var")
+    fi
+done
+
+if [ ${#MISSING_VARS[@]} -gt 0 ]; then
+    echo "❌ Missing required environment variables:"
+    for var in "${MISSING_VARS[@]}"; do
+        echo "   - $var"
+    done
+    echo ""
+    echo "Please configure these in Hugging Face Space Settings > Variables and Secrets"
+    exit 1
+fi
+
+echo "✅ All required environment variables are set"
+echo ""
+
+# Optional variables (log if present)
+echo "📊 Optional services status:"
+[ -n "$COINGECKO_API_KEY" ] && echo "   ✓ CoinGecko API configured" || echo "   ⚠ CoinGecko API not configured (using free tier)"
+[ -n "$GCP_PROJECT_ID" ] && echo "   ✓ Google Cloud SQL configured" || echo "   ⚠ Google Cloud SQL not configured"
+[ -n "$MCP_TOOLBOX_SERVER_URL" ] && echo "   ✓ MCP Toolbox URL: $MCP_TOOLBOX_SERVER_URL" || echo "   ⚠ MCP Toolbox not configured"
+
+# Start MCP Toolbox if Cloud SQL is configured
+if [ -n "$GCP_PROJECT_ID" ] && [ -n "$CLOUD_SQL_INSTANCE" ]; then
+    echo ""
+    echo "🚀 Starting MCP Toolbox (Cloud SQL) on port 5000..."
+    echo "=================================================="
+    
+    # Configure GCP credentials if provided as JSON string
+    if [ -n "$GCP_SERVICE_ACCOUNT_JSON" ]; then
+        echo "   ✓ Configuring GCP credentials from environment variable..."
+        mkdir -p /app/credentials
+        echo "$GCP_SERVICE_ACCOUNT_JSON" > /app/credentials/gcp-key.json
+        export GOOGLE_APPLICATION_CREDENTIALS=/app/credentials/gcp-key.json
+        echo "   ✓ GCP credentials configured"
+    fi
+    
+    # Substitute environment variables in tools.yaml
+    echo "   → Preparing tools.yaml with environment variables..."
+    envsubst < /app/tools.yaml > /tmp/tools.yaml
+    
+    # Start MCP Toolbox server in background
+    echo "   → Starting MCP Toolbox server..."
+    /usr/local/bin/toolbox \
+        --tools-file /tmp/tools.yaml \
+        --address 0.0.0.0 \
+        --port 5000 > /tmp/toolbox.log 2>&1 &
+    
+    TOOLBOX_PID=$!
+    echo "   ✓ MCP Toolbox started (PID: $TOOLBOX_PID)"
+    
+    # Wait for toolbox to be ready
+    echo "   ⏳ Waiting for MCP Toolbox to be ready..."
+    TIMEOUT=30
+    ELAPSED=0
+    while [ $ELAPSED -lt $TIMEOUT ]; do
+        if curl -sf http://localhost:5000/health > /dev/null 2>&1; then
+            echo "   ✅ MCP Toolbox is ready!"
+            break
+        fi
+        sleep 2
+        ELAPSED=$((ELAPSED + 2))
+    done
+    
+    if [ $ELAPSED -ge $TIMEOUT ]; then
+        echo "   ⚠ MCP Toolbox health check timeout (this is OK if no Cloud SQL configured)"
+    fi
+    
+    # Show toolbox logs for debugging
+    echo "   ℹ Check logs: tail /tmp/toolbox.log"
+    if [ -f /tmp/toolbox.log ]; then
+        echo ""
+        echo "📋 MCP Toolbox startup logs:"
+        echo "----------------------------"
+        cat /tmp/toolbox.log
+        echo "----------------------------"
+        echo ""
+    fi
+else
+    echo ""
+    echo "⚠ Skipping MCP Toolbox - Cloud SQL not configured"
+    echo "   Set GCP_PROJECT_ID and CLOUD_SQL_INSTANCE to enable Finance Tracker"
+fi
+
+echo ""
+echo "🎯 Starting Gradio application on port 7860..."
+echo "=================================================="
+echo ""
+
+# Set Python path to include /app
+export PYTHONPATH=/app:$PYTHONPATH
+
+# Debug: Show Python path and verify modules
+echo "🔍 Debug Info:"
+echo "   Working directory: $(pwd)"
+echo "   PYTHONPATH: $PYTHONPATH"
+echo "   Python version: $(python --version)"
+if [ -n "$TOOLBOX_PID" ]; then
+    echo "   MCP Toolbox PID: $TOOLBOX_PID"
+fi
+echo ""
+
+# Launch Gradio app with native MCP support
+cd /app
+exec python -u ui/gradio_app.py