narinzar agents sql mcp

A natural language interface for interacting with SQLite databases, powered by Anthropic's Claude AI.

This application allows users to interact with SQLite databases using natural language. Instead of writing SQL queries manually, users can ask questions or give instructions in plain English, and the AI will generate and execute the appropriate SQL queries, returning formatted results.

• Natural Language Queries: Ask questions about your data in plain English
• SQL Query Generation: Automatically generates SQL queries based on user input
• Query Execution: Executes the generated queries against a connected SQLite database
• Result Formatting: Displays query results in a readable format
• Error Handling: Provides clear explanations when queries fail
• Example Queries: Includes sample queries to help users get started

The application consists of three main components:

1. Web Interface (app.py): A Gradio web interface that allows users to interact with the AI
2. MCP Server (mcp_server.py): A server that handles SQL query execution
3. MCP Client (mcp_client.py): A client that facilitates communication between the web interface and the MCP server

• Python 3.8+
• An Anthropic API key (to access Claude)

1. Clone this repository:

   git clone https://github.com/yourusername/ai-sql-assistant.git
   cd ai-sql-assistant
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Create a .env file in the project root with your Anthropic API key:

   ANTHROPIC_API_KEY=your_api_key_here
   

1. Start the application:

   python app.py
   

2. Open your web browser and navigate to:

   http://localhost:7860
   

3. Type in a natural language query and press Enter or click Submit.

Here are some example queries you can try:

• "List all tables in the database"
• "Show me all users"
• "Count how many products cost more than $300"
• "What's the total value of all products in stock?"
• "Insert a new product called 'Tablet' with price $499.99 and stock 5"
• "Update the price of Laptop to $1099.99"

The application creates a sample SQLite database with the following tables:

• id: INTEGER (Primary Key)
• name: TEXT
• email: TEXT
• created_at: TIMESTAMP

• id: INTEGER (Primary Key)
• name: TEXT
• price: REAL
• stock: INTEGER

The application uses MCP to manage communication between Claude and the SQLite database. MCP allows Claude to call functions (tools) and receive results, creating an interactive feedback loop.

The application uses Anthropic's Claude 3.7 Sonnet model to:

1. Understand natural language queries
2. Generate appropriate SQL queries
3. Explain query results to users

The application uses SQLite for database operations. SQLite is:

• Lightweight
• Serverless
• Self-contained
• Zero-configuration

To use your own SQLite database instead of the sample one:

1. Replace the database initialization in app.py with your own database file path
2. Update the sample table documentation in the UI to match your schema

The system prompt that guides Claude's behavior can be modified in the ChatProcessor class in app.py.

• Currently only supports SQLite databases
• Complex queries might require refinement
• Limited error recovery for malformed queries

Contributions are welcome! Please feel free to submit a Pull Request.

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