This repository contains an MCP (Model Context Protocol) server written in Python that connects to a SQLite database containing startup funding data. The server exposes table schemas as resources, provides a read-only SQL query tool, and offers prompt templates for common data analysis tasks. It is designed to work with MCP clients and language models (LLMs) and communicates via the STDIO protocol.

• Overview
• Features
• Setup and Installation
  • Creating the Sample SQLite Database
  • Creating a Virtual Environment
  • Running the MCP Server
  • Installing into Claude Desktop
• Usage
• Testing
• Logging
• License

The MCP server uses the MCP Python SDK (with CLI extras) to implement a server that:

• Connects to a SQLite database (e.g., a database with startup funding information).
• Exposes table schemas as MCP resources.
• Provides a tool for executing read-only SQL queries.
• Offers prompt templates that help language models generate data analysis insights.
• Communicates via the STDIO protocol, reading JSON-RPC messages from standard input and writing responses to standard output.

• Resources

  • schema://sqlite/{table}: Returns the SQL schema for a specific table.
  • schema://sqlite/all: Returns a JSON mapping of all table schemas.

• Tools

  • sql_query: Executes read-only SQL queries. Only SELECT statements are permitted.

• Prompts

  • analyze_table_prompt: Generates an analysis prompt for a specific table.
  • describe_query_prompt: Generates a prompt explaining a SQL query.

• STDIO Protocol

  • Reads from stdin and writes responses to stdout, making integration easy.

• Logging

  • Uses Python’s logging module to trace activity and debug errors.

Save the following script as create_db.py:

<same as earlier>

Run with:

python create_db.py

python -m venv venv

Activate the environment:

• macOS/Linux:

  source venv/bin/activate

• Windows:

  venv\Scripts\activate

Install dependencies:

pip install "mcp[cli]"

1. Save your server code as sqlite_mcp_server.py.
2. Run the server:

python sqlite_mcp_server.py

Optional (using uv):

uv run sqlite_mcp_server.py

Save the following as install_to_claude.py and run it:

python install_to_claude.py

Update Claude Desktop config with:

{
  "mcpServers": {
    "sqlite_mcp_server": {
      "command": "python",
      "args": ["-u", "/absolute/path/to/sqlite_mcp_server.py"]
    }
  }
}

Restart Claude Desktop afterward.

• Access Resources: schema://sqlite/all, schema://sqlite/startups
• Invoke Tools:

  SELECT * FROM startups WHERE funding_amount > 10000000;

• Use Prompts: Generate SQL or explain queries.

Run the test script:

python sqlite_mcp_client_tests.py sqlite_mcp_server.py

This script tests:

• Listing resources
• Retrieving schemas
• Valid/invalid queries
• Prompt templates

Modify logging config:

logging.basicConfig(
    filename='mcp_server.log',
    level=logging.DEBUG,
    format="[%(asctime)s] %(levelname)s - %(message)s",
    datefmt="%Y-%m-%d %H:%M:%S"
)

Logs print to stderr by default.

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

• Adjust paths in installation/config scripts as needed.
• Ensure the SQLite database is created before running the server.
• Integrates with MCP Inspector or Claude Desktop.