MCP-Based Production AI Agent Platform

A production-grade FastAPI application with MCP (Model Context Protocol) architecture featuring RBAC, streaming, real-time data analysis, and multi-database integration.

🏗️ Architecture

Modular MCP Server Design:

• MongoDB Server (Port 8100): File storage, metadata, S3 integration
• Milvus Server (Port 8110): Vector database for embeddings
• WebSearch Server (Port 8140): Real-time web search capabilities
• SciREX Server (Port 8150): ML analysis and data processing

Core Components:

• FastAPI Production App: RBAC, streaming, error handling
• MCP Agent: Dynamic tool discovery with role-based access
• Multi-Database: MongoDB, Redis, Milvus integration
• Real-time Analysis: Immediate data profiling and metadata enrichment

🚀 Quick Start

Essential Commands

make pre_start      # Complete setup: environment, dependencies, Docker & MCP servers
make run_app        # Start Streamlit web application  
make stop          # Stop all services and cleanup

Manual Setup

# 1. Environment setup
cp .env.sample .env  # Configure your settings

# 2. Start services
docker compose up -d

# 3. Test connectivity  
python3 tests/connectivity_test.py

# 4. Run application
streamlit run ui_streamlit.py

📁 Project Structure

mcp/
├── agent/          # MCP Agent core logic with RBAC
├── api/            # FastAPI production endpoints  
├── config/         # Configuration management
├── mcp_client/     # MCP client implementation
├── mcp_servers/    # MCP server implementations
│   ├── mongo_server/     # MongoDB + S3 storage
│   ├── milvus_server/    # Vector database
│   ├── websearch_server/ # Web search capabilities
│   └── scirex_server/    # ML analysis server
├── utils/          # Shared utilities
├── tests/          # Test suite
└── docs/           # Documentation

🧪 Testing

# Health check
python3 tests/connectivity_test.py

# File upload workflow
python3 tests/test_direct_upload.py

# Full test suite
pytest tests/

📊 Features

• ✅ Production FastAPI with RBAC, streaming, error handling
• ✅ MCP Protocol compliance with dynamic tool discovery
• ✅ Multi-database support (MongoDB, Redis, Milvus)
• ✅ S3 Storage integration with automatic file classification
• ✅ Real-time Analysis with immediate metadata enrichment
• ✅ User Context propagation and role-based access control

📚 Documentation

See docs/ folder for detailed documentation:

• Architecture and implementation guides
• Development history and fixes
• Test workflow procedures

---

Status: Production-ready MCP Agent system with full database connectivity and file upload workflow. ✅