READMEmain

MCP Streamlit Chatbot

This repository is meant as a hands-on practice of working with MCP in a Streamlit chatbot that is meant to be deployed in AWS.

Project Structure

mcp-streamlit-chatbot/
├── app/                    # Streamlit application
│   ├── main.py            # Main Streamlit application file
│   ├── mcp_client.py      # MCP client implementation
│   └── tools/             # Application-specific tools
├── data/                  # Data directory for PDFs and other files
│   └──
├── docker/                # Docker configuration files
│   ├── docker-compose.yml # Container orchestration
│   ├── Dockerfile.app     # Streamlit app container
│   └── Dockerfile.mcp     # MCP server container
├── mcp-server/           # MCP server implementation
│   ├── main.py           # MCP server entry point
│   └── tools/            # MCP server tools
│       └── summarize_pdf.py
├── Makefile              # Build and deployment commands
└── requirements.txt      # Python dependencies

Prerequisites

Docker and Docker Compose
Make (optional, for using Makefile commands)

Setup and Running

Clone the repository:

git clone <repository-url>
cd mcp-streamlit-chatbot

Build and start the containers:
```
make build
make up
```
Access the applications:
- Streamlit App: http://localhost:8501
- MCP Server: http://localhost:8000

Available Make Commands

The project includes a Makefile with the following commands:

make help      # Show all available commands
make build     # Build the containers
make up        # Start containers in detached mode
make down      # Stop and remove containers
make ps        # List running containers
make logs      # Show logs from all containers
make up-logs   # Start containers in foreground mode (with logs)
make clean     # Remove all containers, images, and volumes
make rebuild   # Rebuild and restart containers
make restart   # Restart containers

Service-Specific Commands

# Show logs for a specific service
make service-logs SERVICE=streamlit-app

# Execute a command in a container
make exec SERVICE=streamlit-app CMD=bash

Development

The application consists of two main components:

Streamlit App (app/):
- Provides the chat interface
- Communicates with the MCP server
- Handles user interactions
MCP Server (mcp-server/):
- Implements the MCP protocol
- Provides tools for PDF processing
- Handles backend operations

Data Directory

The data/ directory is mounted to both containers:

Streamlit App: /app/data
MCP Server: /mcp/data

Place any PDFs or other files you want to process in this directory.

Contributing

Fork the repository
Create a feature branch
Commit your changes
Push to the branch
Create a Pull Request

mcp-streamlit-chatbot