Langgraph-MCP Client for Postgresql Example

This repository demonstrates a very low level example of how to interact with an MCP server as a docker image, connected to a PostgreSQL database for LangGraph agents.

model context protocol

MCP is an open protocol that standardizes how applications provide context to LLMs. Think of MCP like a USB-C port for AI applications. Just as USB-C provides a standardized way to connect your devices to various peripherals and accessories, MCP provides a standardized way to connect AI models to different data sources and tools.

Prerequisites

Before you begin, ensure you have the following installed on your system:

• Docker
• PostgreSQL Database - running and populated
• MCP Servers

Flow and tools

• We use a postgresql database that the Postgresl MCP Server is aware of through the config
• from langchain_mcp_adapters.tools import load_mcp_tools creates the mcp tool
• create_react_agent is a fast agent templater
• we output the stream to file

Project Setup Instructions

1. Clone the Repository

Clone this repository to your local machine:

git clone https://github.com/your-repo/Langchain-MCP.git
cd Langgraph-mcp-client directory

2. Create a .env File

Create a .env file in the root of the repository with the following variables:

DB_HOST=<your-database-host>
DB_PORT=<your-database-port>
DB_USER=<your-database-username>
DB_PASSWORD=<your-database-password>
DB_NAME=<your-database-name>

Replace the placeholders with your actual database and MCP server connection details.

3. Install UV Package

Install the UV package globally if you haven't already:

Astral uv

MCP Server Setup

1. Clone the repo MCP Servers to a separate directory
2. Run - docker build -t mcp/postgres -f src/postgres/Dockerfile . to build and tag the image

Usage Instructions

1. Sync the UV Package

Run the following command to sync the UV package:

uv sync

This command ensures that the UV package installs the package correctly.

2. Run the Query Agent

Start the query agent using the following command:

uv run queryagent

This will prompt for the user query

Alt - with query (Banking database)

uv run queryagent "Who holds the most funds in thier account?"

TODO:

• Output formating
• Multi agent

Troubleshooting

• Database Connection: Verify the .env file contains the correct database credentials.
• UV Package Errors: Ensure the UV CLI is installed globally and the uv sync command completes without errors.

Contributing

Feel free to open issues or submit pull requests to improve this repository.

License

This project is licensed under the MIT License. See the LICENSE file for details.