sqlite-explorer-fastmcp

This MCP server provides safe, read-only access to a SQLite database via the Model Context Protocol. It exposes tools that let an LLM explore the schema and execute read-only queries against the database with built-in safety validations, parameter binding, and row-limiting to prevent large or unsafe results. Key tools include read_query for executing SELECT statements with validation, list_tables to enumerate available tables, and describe_table to fetch detailed schema information such as column types, nullability, defaults, and primary keys. To use the server, point your client or environment to the uv-based runtime (as shown in the installation instructions) and supply the path to your SQLite database via the SQLITE_DB_PATH environment variable. The server will enforce read-only access and sanitize queries to prevent unsafe operations.

Prerequisites:

• Python 3.6+
• Git (optional, for cloning)
• Access to the internet to install dependencies

Option A: Install and run locally

1. Clone or download the repository containing sqlite_explorer.py and requirements.txt
2. Create and activate a Python environment (recommended):
   • python3 -m venv venv
   • source venv/bin/activate (Linux/macOS) or venv\Scripts\activate (Windows)
3. Install dependencies:
   • pip install -r requirements.txt
4. Set up your database path (example):
   • export SQLITE_DB_PATH=/path/to/your/database.db
5. Run the MCP server using the provided uv-based command:
   • uv run --with fastmcp --with uvicorn fastmcp run sqlite_explorer.py
   • Ensure SQLITE_DB_PATH is set in the environment before starting (the server will read this variable at startup)

Option B: Use Claude Desktop or Cline VSCode plugin (as described in the README)

1. Ensure Python and dependencies are installed as in Option A
2. Start the server with the appropriate environment variable and command configuration as shown in the README example.

Environment variables:

• SQLITE_DB_PATH: Full path to the SQLite database file to explore. Safety features:
• Read-only access to the database to prevent modifications.
• Query validation and sanitization with parameter binding.
• Row limit enforcement to avoid large payloads.
• Progress output suppression to ensure clean JSON responses. Common issues:
• Ensure SQLITE_DB_PATH is accessible by the user running the server.
• If the server fails to start, verify that Python 3.6+ is installed and dependencies from requirements.txt are installed.
• When using the UVicorn-based runtime, ensure that the uv command is available in your PATH.