What is Mcp Turso

mcp-turso is a Model Context Protocol (MCP) server designed for interacting with LibSQL databases hosted on Turso. It facilitates various database operations such as retrieving table lists, database schemas, and executing SELECT queries.

Use cases

mcp-turso can be used in scenarios where applications need to interact with LibSQL databases, such as data retrieval for analytics, application backend integrations, and database management tasks.

How to use

To use mcp-turso, configure it in your ‘claude_desktop_config.json’ file by specifying the command, arguments, and environment variables for your Turso database URL and authentication token. Ensure you have an existing database and retrieve the necessary credentials using the Turso CLI.

Key features

Key features of mcp-turso include: retrieving a list of tables, obtaining the database schema, describing specific table schemas, and executing SELECT queries to fetch data.

Where to use

undefined

mcp-turso

A Model Context Protocol (MCP) server that provides access to the Turso-hosted LibSQL databases. Currently, the server provides the following functionality:

• Retrieving a list of tables in a database
• Retrieving a database’s schema
• Retrieving the schema of a table
• Performing SELECT queries

Configuration

With Claude Desktop

Add this to your claude_desktop_config.json:

You will need an existing database to continue. If you don’t have one, create one. To get the database URL via the Turso CLI, run:

turso db show --url <database-name>

Then get the database authentication token:

turso db tokens create <database-name>

Add those values to your configuration as shown above.

With Cursor

To configure the Turso MCP server with Cursor, add the following to your Cursor settings:

1. Open Cursor and go to Settings (⚙️) > Settings (JSON)
2. Add the following configuration to your settings JSON:

Replace your_url and your_token with your Turso database URL and authentication token as described in the previous section.

Logging

The server includes a custom logger for debugging outside of Claude Desktop. By default, this logger writes to <parent-dir>/logs/mcp-turso.log, where <parent-dir> is the parent directory of directory containing the mcp-turso script. In other words, if the path to mcp-turso is ~/foo/bin/mcp-turso, the logs will be at ~/foo/logs/mcp-turso.log. If running with NPX as above, the default logs will be:

~/.npm/_npx/<npx-dir-name>/node_modules/mcp-turso/logs/mcp-turso.log

If you would like to specify a custom path, you can include a --logs flag with an absolute posix path in the server’s configuration:

The path to the log file (default or custom) is always logged to stderr when the server is created. For Claude desktop, this will show up in your server logs in ~/Library/Logs/Claude.

Note: Right now, I haven’t implemented specifying a custom logging file for Windows, but this is coming.

Server Capabilities

The server provides the following tools:

• list_tables
  • Get a list of all the tables in the database
  • No input
  • Returns: an array of table names
• get_db_schema
  • Get the schemas of all tables in the database
  • No input
  • Returns: an array of SQL creation statements
• describe_table
  • View schema information for a specific table
  • Input:
    • table_name (string): Name of table to describe
  • Returns: Array of column definitions with names and types
• query_database
  • Execute a SELECT query to read data from the database
  • Input:
    • sql (string): The SELECT SQL query to execute
  • Returns: Query results as an object of type { columns: string[]; rows: Record<string, unknown>[]; rowCount: number; }

Todo

• [ ] Add the ability to specify a custom log file on windows
• [ ] Add more query tools

License

MIT License - see the LICENSE file for details.