Which integrations are available for this server?

Provided as a method to test the MCP server's API endpoints for tool discovery and execution. Provides configuration management through environment variables loaded from a .env file for database credentials and other settings. Used as the web framework for building the MCP server, handling HTTP requests and routing for tool discovery and execution. Used for version control and distribution of the MCP server codebase. The programming language used to implement the MCP server and its tools.

How do I use Modular MCP Server?

1. Click on "Install Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Modular MCP Server what tools are currently available?" That's it! The server will respond to your query, and you can continue using it as needed. Here is a step-by-step guide with screenshots.

Modular MCP Server

This is a modular, auto-discovering Model Context Protocol (MCP) server built with Python and FastAPI. It's designed to be a scalable foundation for providing tools to Large Language Models (LLMs).

Features

Auto-Discovery: Automatically loads tools from the /tools directory on startup.
Modular: Tools are self-contained in their own files.
Configurable: Uses a .env file for all configuration.
Secure: Includes best practices like read-only database users and whitelisting.
Performant: Runs on a high-performance Unix socket and uses an async database connection pool.

Related MCP server: MCP Web Tools Server

Setup

Clone the repository:
```
git clone <repository_url>
cd mcp
```

Create and activate a virtual environment:

python3 -m venv .venv
source .venv/bin/activate

Install dependencies:
```
pip install -r requirements.txt
```
Configure your environment:
```
cp .env.example .env
```
Now, edit the .env file with your actual database credentials and settings.

Running the Server

To run the server on a Unix socket:

uvicorn mcp_server:app --uds /tmp/mcp.sock

Testing

You can test the running server using curl.

Discover tools:

curl --unix-socket /tmp/mcp.sock http://localhost/mcp/v1/tools | jq

Execute a tool:

curl -X POST \
  --unix-socket /tmp/mcp.sock \
  -H "Content-Type: application/json" \
  -d '{"arguments": {"timezone": "Europe/Istanbul"}}' \
  http://localhost/mcp/v1/tools/get_current_time:execute | jq