This server wraps Python's sqlite3 library and exposes it through MCP tools. You get connection management (connect, close, commit, rollback), direct SQL execution (execute_query, execute_many, execute_script), schema inspection (list_tables, get_table_info), and convenience CRUD operations for common database tasks. Reach for this when you want Claude to interact with SQLite databases without writing custom integration code. It handles the full lifecycle from opening connections to running queries to managing transactions. The tool set is comprehensive enough for both quick data queries and more complex database operations.
MCP server exposing sqlite3 library functionality
pip install mcp-sqlite3
from mcp_sqlite3 import mcp_server
mcp_server.run()
mcp-sqlite3 --help
Exposes complete sqlite3 functionality as MCP tools.
connect_database - Open a database connectionclose_connection - Close a database connectioncommit / rollback - Transaction controlexecute_query - Execute a query and return resultsexecute_many - Execute with multiple parameter setsexecute_script - Execute a SQL scriptlist_tables - List all tablesget_table_info - Get table schemacreate_table / drop_table - DDL operationsinsert_row / update_rows / delete_rows / select_rowsgit clone https://github.com/daedalus/mcp-sqlite3.git
cd mcp-sqlite3
pip install -e ".[test]"
# run tests
pytest
# format
ruff format src/ tests/
# lint
ruff check src/ tests/
# type check
mypy src/
mcp-name: io.github.daedalus/mcp-sqlite3